US Pat. No. 10,970,703

SECURE ELECTRONIC PAYMENT SYSTEM

PAAY LLC, New York, NY (...

1. A method for undertaking a purchase transaction between a customer and a merchant comprising:receiving, by an authentication notification server, from a merchant, a request to permit transfer of funds from a funds transfer account associated with a customer to the merchant,
wherein the request comprises identifying information that does not comprise an account number associated with the funds transfer account;
correlating, by the authentication notification server, the identifying information to a token received from a token generation and information storage server,
wherein the token corresponds to the funds transfer account, and
wherein the token does not comprise the account number associated with the funds transfer account;
upon correlating the identifying information to the token, transmitting, by the authentication notification server, the token to the token generation and information storage server to thereby cause a request for utilization of the funds transfer account to be transmitted;
receiving, by the authentication notification server, an approval relating to the request for utilization of the funds transfer account; and
transferring, by the authentication notification server, the approval to the merchant.

US Pat. No. 10,970,702

SYSTEMS AND METHODS FOR FACILITATING MULTI-PARTY PAYMENT TRANSACTIONS

MASTERCARD INTERNATIONAL ...

1. A computer-implemented method for enabling communications through the Internet between a plurality of cardholders and a separate buyer for a purchase of an item by one of the plurality of cardholders on behalf of the buyer, the method implemented by a host computing device coupled to a memory, said method comprising:hosting, by the host computing device, a transaction portal communicatively coupled, via the Internet, to a plurality of cardholder computing devices and a buyer computing device, wherein each of the plurality of cardholder computing devices is associated with a respective cardholder of the plurality of cardholders, wherein each of the plurality of cardholders is associated with a respective cardholder account, and wherein the buyer computing device is associated with the buyer;
receiving, by the host computing device via the transaction portal, cardholder data associated with each respective cardholder account, the cardholder data including at least one merchant association representing a merchant payment card issued to the respective cardholder;
receiving, by the host computing device via the transaction portal, proposal data from the buyer computing device, wherein the proposal data identifies the item proposed for purchase, a particular merchant associated with a particular merchant payment card, and a merchant purchase price;
filtering, by the host computing device using the proposal data and the cardholder data, a subset of relevant cardholders from the plurality of cardholders, wherein each cardholder of the subset of relevant cardholders includes the particular merchant payment card associated with the particular merchant identified in the proposal data;
causing, by the host computing device via the transaction portal, a list of the subset of relevant cardholders to be displayed on a user interface of the buyer computing device;
receiving, by the host computing device via the transaction portal, a selection message from the buyer computing device, wherein the selection message indicates a selected cardholder of the subset of relevant cardholders chosen to purchase the item on behalf of the buyer using the particular merchant payment card associated with the particular merchant;
reserving, by the host computing device, at least the merchant purchase price and a cardholder processing fee within a buyer account associated with the buyer; and
transferring, by the host computing device, funds from the buyer account to a selected cardholder account associated with the selected cardholder, the funds determined at least in part by the merchant purchase price and the cardholder processing fee.

US Pat. No. 10,970,701

SYSTEM FOR IDENTIFYING OR ASSISTING THE IDENTIFICATION OF A PRODUCT OR SET OF PRODUCTS

JES Labs, Saint Herblain...

1. An identification or assistance in identification system for identifying or assisting in identifying a product or a set of products of a user, said identification or assistance in identification system comprising:a counter for receiving said product or set of products;
a lighting system for lighting the counter;
a plurality of image acquisition devices pointing towards the counter so as to acquire images of said product or set of products at different viewing angles; and
a display screen;
the identification or assistance in identification system being wherein said system further comprises:
a first database storing user account data, each user account comprising:
an identifier of a user;
a credit or debit balance value;
an identifier device enabling the user to be identified; and
a control device configured:
to acquire an identifier of the user;
to determine whether the acquired identifier of the user corresponds to the identifier of the user of a stored user account;
to acquire images of said product or set of products by using said image acquisition devices; and
to associate the acquired images with the identifier of the user in a second database,
wherein the identification or assistance in identification system further comprises a back face that extends substantially orthogonally to the counter, the back face absorbing daylight less than the counter; and
wherein said image acquisition devices include at least a first image acquisition device having its optical axis oriented perpendicularly to the counter, and second and third image acquisition devices arranged on either side of a plane containing the optical axis of the first image acquisition device and perpendicular to the back face, the optical axes pointing towards the counter and sloping in mutually opposite directions relative to the optical axis of the first image acquisition device.

US Pat. No. 10,970,700

INFORMATION PROCESSING APPARATUS AND SECURITY CONTROL METHOD

TOSHIBA TEC KABUSHIKI KAI...

1. An information processing apparatus which is removably mountable on a docking station, the information processing apparatus comprising:an input device to receive an operation input;
an acquisition unit to acquire information indicating an operator; and
a processor configured to:
detect a removed state in which the information processing apparatus is removed from the docking station,
upon detecting the removed state, control the acquisition unit to start an operation to acquire the information indicating the operator and perform user authentication based on the information indicating the operator acquired through the operation of the acquisition unit,
when the operator indicated in the information acquired through the operation of the acquisition unit is approved as an authorized operator of the information processing apparatus, based on a result of the user authentication, control the input device to enable the operation input, and
when the operator is not approved as the authorized operator of the information processing apparatus, based on a result of the user authentication, control the input device to disable the operation input until the information processing apparatus is mounted on the docking station.

US Pat. No. 10,970,699

POINT OF SALE PAIRING TO WIRELESS NETWORKS

1. Network equipment, comprising:a processor; and
a memory that stores instructions that, when executed by the processor, facilitates performance of operations, comprising:
receiving an access notification from a router that establishes wireless connectivity to a network, the access notification specifying a user device identifier associated with a user device requesting access to the network;
determining a security credential based on querying a data store for the user device identifier, wherein the data store comprises associations of groups of identifiers to security credentials comprising the security credential that is associated with the user device identifier of the groups of identifiers, wherein the groups of identifiers comprise stock keeping unit data associated with the user device and international mobile subscriber identity data associated with the user device; and
sending a proxy instruction to the router, the proxy instruction specifying the security credential and instructing the router to accept the security credential presented as a proxy on behalf of the user device requesting the access to the network.

US Pat. No. 10,970,698

READER DETECTION SIGNAL BYPASSING SECURE PROCESSOR

Square, Inc., San Franci...

1. A transaction object reader device that provides feedback to a user based on proximity of a transaction object to a transaction object read interface, the transaction object reader device comprising:the transaction object read interface that reads transaction object information from the transaction object in response to physical proximity between the transaction object and the transaction object read interface;
a filter that outputs a filtered signal to a first processor in response to receipt of the transaction object information at the filter from the transaction object read interface, wherein transaction information in the transaction object information is at least partially attenuated out in the filtered signal;the first processor, wherein the first processor receives the filtered signal from the filter, generates the feedback based on the filtered signal, and outputs the feedback to an output device, wherein the feedback indicates how close the transaction object is to the transaction object read interface;the output device, wherein the output device outputs the feedback to the user in response to receipt of the feedback from the first processor; and
a second processor that receives the transaction object information from the transaction object read interface, that generates a processed dataset using the transaction object information, and that outputs the processed dataset, wherein the second processor outputs the processed dataset after the filter outputs the filtered signal.

US Pat. No. 10,970,697

TRANSACTION MEDIATION METHOD

Poynt Co., Palo Alto, CA...

1. A method comprising:with an intermediary system comprising a set of physical processors:
receiving a transaction request comprising transaction information and payment information, wherein the payment information is dynamically issued by a DPP (digital payment platform), separate from the intermediary system, to a user in response to a payment information request, and wherein the transaction information comprises a transaction amount;
after receiving the transaction request, determining whether a DPP (digital payment platform) user account associated with the payment information has sufficient funds for the transaction request;
responsive to a determination that the DPP user account has sufficient funds for the transaction request, generating a temporary payment card number, different from the payment information, that uses the DPP as a funding source;
generating a payment request based on the transaction information and the temporary payment card number; and
transmitting the payment request to a payment network, separate from the DPP, wherein the payment request is a request for requesting payment from the temporary payment card number to a merchant account associated with the transaction request.

US Pat. No. 10,970,696

TECHNIQUES TO PROVIDE PHYSICAL TRANSACTION CARD CAPABILITIES FOR A MOBILE DEVICE

Capital One Services, LLC...

1. A user device, comprising:a near-field communication (NFC) device configured to detect NFC devices to perform transactions;
a housing comprising a front portion, a back portion, and edges to couple the front portion with the back portion;
a substrate coupled with the housing, the substrate comprising a magnetic stripe, and the magnetic stripe programmable to store information associated with at least one account of a plurality of accounts;
substrate controller circuitry located within the housing and coupled with the substrate, the substrate controller circuitry operable to execute instructions to control the substrate, the substrate controller circuitry, when executing the instruction, to:
detect an event to cause at least a portion of the substrate to extend from an edge of the housing, wherein the portion of the substrate comprises the magnetic stripe, wherein the event comprises a failed attempt to perform a transaction via the NFC device with one of the NFC devices;
detect a touch input via a touchscreen display, and the touch input comprises a swipe in a direction toward the edge from which the portion of the substrate extends;
determine a rate of the swipe of the touch input; and
in response to the event, automatically cause the portion of the substrate to extend from the edge of the housing at a second rate proportional to the rate of the touch input.

US Pat. No. 10,970,695

SECURE REAL-TIME TRANSACTIONS

EARLY WARNING SERVICES, L...

1. A system comprising:one or more processors; and
one or more non-transitory computer-readable media storing computing instructions configured to run on the one or more processors and perform:
receiving, at a mobile device being used by a consumer, a transaction code at a mobile device from a point-of-sale terminal, wherein:
the mobile device is used by a consumer;
the point-of-sale terminal is at a store of a merchant and the transaction code comprises (1) a merchant public identifier associated with the merchant, (2) a payment amount for one or more items the consumer is purchasing from the merchant by the consumer in a transaction, (3) a transaction identifier for the transaction, and (4) a digital signature of the transaction code;
verifying the digital signature for the transaction code; and
sending, from the mobile device, a first request from the mobile device to a consumer financial institution, wherein:
the first request is a request to pay the merchant for the payment amount from a consumer account of the consumer maintained by a first the consumer financial institution;
the first request comprises the merchant public identifier, the payment amount, and the transaction identifier;
sending the first request to the consumer financial institution causes:
(1) the first consumer financial institution, upon receiving the first request to retrieve from a transaction system a merchant account identifier of a merchant account of the merchant maintained by a merchant financial institution, wherein the transaction system determines the merchant account identifier to return to the consumer financial institution based on the merchant public identifier;
(2) the consumer financial institution to send to the transaction system payment information regarding a payment to be made to the merchant account from the first consumer account, wherein the payment information comprises (a) the transaction identifier, (b) the merchant account identifier of the merchant account, and (c) the payment amount;
(3) the payment information to be routed from the consumer financial institution through the transaction system to the second merchant financial institution; and
(4) the point-of-sale terminal to receive a notification of good funds for the transaction in real-time while the consumer remains at the store of the merchant; wherein
the transaction system is maintained by an entity that is different from the merchant, the consumer financial institution, and the merchant financial institution;
the merchant does not receive any consumer account identifiers of the consumer account of the consumer; and
the consumer does not receive the merchant account identifier of the merchant account of the merchant.

US Pat. No. 10,970,694

SYSTEM FOR INTERACTIVE MARKETING IN RECONFIGURABLE VENDOR SPACE

WASHINGTON PRIME GROUP L....

1. A system, comprising:a communication gateway within a shared commercial region, the gateway communicatively coupled to a local network and at least one host server, the host server providing an interface for presenting information relating to one or more tenant vendors of the shared commercial region;
a plurality of interaction terminals communicatively coupled to the communication gateway through the local network, each of the plurality of terminals configured to allow interaction with one or more websites hosted by at least one of the tenant vendors of the shared commercial region, wherein:
at least one electronic display of the plurality of interaction terminals is remotely reconfigurable by at least one of the tenant vendors via a web server,
at least one layout of the at least one electronic display is controlled by at least one remote host server in communication with the web server, and
the plurality of interaction terminals are configured to interface with a website operated by an entity distinct from the one or more tenant vendors and configured to re-route to at least one website operated by the one or more tenant vendors; said system further comprising
a common area of the shared commercial region, wherein:
the common area is configured to permit pedestrian travel among the plurality of interaction terminals; and
the common area is configured to present one or more physical representations of goods offered by a vendor tenant, the one or more physical representations located within the shared commercial space;
wherein each of the interaction terminals is configured to:
present product information from a selected tenant vendor; and
invite a customer to interact with the one or more physical representations of goods corresponding to the selected tenant vendor's product information,
wherein the system comprises a redeployable system that is configured to be disassembled at an original location and reassembled at a new location.

US Pat. No. 10,970,693

SEMI-AUTOMATIC CONFIGURATION OF A SELF-SERVICE TERMINAL

NCR Corporation, Atlanta...

1. A method, comprising:providing executable instructions to a processor of a Self-Service Terminal (SST) from a non-transitory computer-readable storage medium causing the processor to perform operations comprising:
obtaining configuration connection details by an input device of the SST, wherein obtaining further includes obtaining, by the input device of the SST, the configuration connection details directly from a second device that is independent of the SST, wherein the input device is a Near Field Communications (NFC) receiver;
identifying a location from the configuration connection details that identifies for the SST a configuration server for obtaining configuration information to configure the SST, wherein the location is external to the SST;
connecting the configuration server at the location to the SST by processing the configuration connection details;
obtaining the configuration information from the configuration server while the SST is connected to the configuration server;
processing the configuration information that configures the SST with the configuration information including configuring the SST for interaction with a local server and an external system and providing by processing the executable instructions a self-configuration of the SST; and
performing transactions on the SST utilizing the configuration information.

US Pat. No. 10,970,692

METHOD, SYSTEM AND SERVER SYSTEM OF PAYMENT BASED ON A CONVERSATION GROUP

TENCENT TECHNOLOGY (SHENZ...

1. A method of performing a payment transaction in a group chat environment, comprising:at a user device having one or more processors and memory for storing one or more programs to be executed by the one or more processors:
displaying a conversation interface of a group chat session for a pre-established conversation group, the pre-established conversation group including a user of the user device and a plurality of other users participating in the pre-established conversation group;
providing, in the conversation interface of the group chat session, a control for requesting payment from the pre-established conversation group;
detecting an input from the user invoking the control for requesting payment from the pre-established conversation group;
in response to detecting the input from the user:
posting a group payment request as a chat message in the group chat session to one or more designated payers in the pre-established conversation group,
wherein the one or more designated payers for the group payment request are determined in accordance with a predetermined payment scheme associated with the pre-established conversation group, wherein the predetermined payment scheme specifies one or more criteria for dynamically selecting a subset of users from the pre-established conversation group as the one or more designated payers for the group payment request, and wherein the group payment request is only visible and accessible to the one or more designated payers; and
receiving a plurality of payment responses as chat messages in the group chat session from the determined designated payers, each payment response including an indication of accepting or rejecting the group payment request;
in accordance with a determination that a first number of the determined designated payers have accepted the group payment request, obtaining a first distribution of payments among the first number of the determined designated payers according to the predetermined payment scheme; and
sending the first distribution of payments to a server, wherein the server transfers the payments from the first number of the determined designated payers to the user of the user device.

US Pat. No. 10,970,691

SYSTEM, METHOD, AND APPARATUS FOR REPROGRAMMING A TRANSACTION CARD

CAPITAL ONE SERVICES, LLC...

1. A user device comprising:one or more processors;
an antenna;
an input/output interface comprising an embedded near-field communication (NFC) chip; and
memory configured to store instructions to cause the one or more processors to:
receive, at the input/output interface, one or more authentication factors from a user;
determine, with the one or more processors, that the one or more authentication factors match one or more stored authentication factors;
receive new transaction card data from an account provider system;
create, with the antenna, a secure wireless connection between the user device and a reprogrammable transaction card;
transmit, via the secure wireless connection, instructions to the reprogrammable transaction card to delete at least a portion of existing transaction card data securely stored on an embedded microchip of the reprogrammable transaction card;
transmit, via the secure wireless connection, the new transaction card data and instructions to store the new transaction card data on the embedded microchip;
receive, via the secure wireless connection, a response from the reprogrammable transaction card;
determine, with the one or more processors, that the new transaction card data is stored on the reprogrammable transaction card based on the response;
deactivate, with the NFC chip of the input/output interface and the one or more processors, the secure wireless connection between the user device and the reprogrammable transaction card; and
transmit a confirmation to the account provider system that the reprogrammable transaction card has been updated with the new transaction card data.

US Pat. No. 10,970,690

BLOCKCHAIN DISTRIBUTION NETWORK WITH SERVICE TRANSPARENCY

Northwestern University, ...

1. A peer node in a blockchain distribution network, the peer node comprising:a processor configured to generate an encrypted test block, wherein the encrypted test block is indistinguishable from an encrypted blockchain in the blockchain distribution network; and
a transceiver operatively coupled to the processor and configured to:
transmit the encrypted test block to a server for propagation throughout the blockchain distribution network; and
receive a message from a second peer node in the blockchain distribution network, wherein the message indicates that the second peer node received the encrypted test block;
wherein the processor is configured to determine an elapsed time between transmission of the encrypted test block and receipt of the message, and wherein the processor is configured to determine whether the peer node is being discriminated against with respect to blockchain transmissions based at least in part on the elapsed time.

US Pat. No. 10,970,689

SYSTEMS AND METHODS FOR DYNAMIC USER INTERFACE MODIFICATION

The Toronto-Dominion Bank...

9. A method of conditionally altering speed of transactions on a computing device having an input device and a display, the computing device being configured to communicate with a remote server over a data network, the method comprising:receiving a digital ID from the input device;
determining that the digital ID is associated with one or more accounts including at least one credit account;
receiving, from the input device, a request to initiate a data exchange with the at least one credit account; and,
determining whether the data exchange, when implemented, would result in the at least one credit account exceeding a utilization threshold and,
when the utilization threshold would not be exceeded:
displaying an interface on the display to enable completion of the data exchange in a first period of time, the interface including an input field; and
pre-populating the input field with a default entry;
when the utilization threshold would be exceeded:
displaying a modified interface different from the interface to enable completion of the data exchange in a second period of time longer than the first period of time, the modified interface including the input field without a pre-populated default entry; and
receiving, via the input device, an active input with respect to the input field.

US Pat. No. 10,970,688

SYSTEM AND METHOD FOR TRANSFERRING FUNDS

EARLY WARNING SERVICES, L...

1. A method comprising:registering, by a computer system of a second financial institution, a recipient using at least one public identifier, each of the at least one public identifier comprising contact information to contact the recipient, the recipient being registered by creating at least one record in an information directory at the second financial institution, the at least one record comprising a private identifier of the recipient, and the private identifier being usable by the second financial institution to determine an account number of a recipient account of the recipient at the second financial institution;
receiving an expedited funds transfer message, the expedited funds transfer message requesting that the second financial institution provide recipient funds to the recipient account at the second financial institution before sender funds are received by the second financial institution from a first financial institution, the first and second financial institutions being members of a computer-implemented payment network, and the expedited funds transfer message comprising the private identifier;
determining the account number of the recipient account using the private identifier;
before the second financial institution receives the sender funds, the second financial institution providing the recipient funds to the recipient account using the account number of the recipient account before the second financial institution receives the sender funds, the recipient funds being provided to the recipient account after (i) the sender funds were removed from a sender financial account by the first financial institution and (ii) the expedited funds transfer message was sent by the first financial institution;
notifying the recipient, using at least one of the at least one public identifier, that a payment has been provided to the recipient account in real-time after initiation of the expedited funds transfer message from the first financial institution,
wherein:
the sender funds are received at the second financial institution from the first financial institution after the recipient funds are provided to the recipient account.

US Pat. No. 10,970,687

ROUTING TRANSACTION DATA OVER A DATA PIPE

Motorola Mobility LLC, C...

1. A method comprising:detecting, at a first device, initiation of a financial transaction with a second device, wherein the financial transaction is initiated to be remotely processed by a transaction service, wherein the first device is a payer device connected to the transaction service via a primary connection which meets a minimum connection service criteria, and wherein the second device is a payee device, which defaults to complete the financial transaction by establishing and communicating second device transaction data over a second connection with the transaction service when the second connection to the transaction service also meets the minimum connection service criteria;
in response to receiving, from the second device, a notification indicating that the second device does not have a connection meeting the minimum connection service criteria for connecting to the transaction service to complete an exchange of the second device transaction data of the financial transaction, establishing, utilizing a wireless radio of the first device having a transceiver for exchanging data with a connected wireless radio of the second device, a data pipe that directly connects the first device and the second device to enable transfer of the second device transaction data between the first and second devices; and
in response to the data pipe being established, the first device: receiving, over the data pipe, the second device transaction data intended for forwarding to the transaction service from the second device, the second device transaction data being generated by a financial transaction client on the second device to provide access via an authenticated login to resources of a client, account, or user associated with the second device; and automatically routing, via the primary connection to the transaction service, the second device transaction data received over the data pipe from the second device, the second device transaction data comprising information that identifies a particular transaction service to be used to process the financial transaction, communication routing information for accessing the transaction service, and information that authorizes the transaction service to execute the financial transaction;
wherein: the minimum connection service criteria identifies, for connections to the transaction service, at least one of at least one permitted type of connection and at least one restricted type of connection, and a minimum connection security level between a respective one of the first and the second device and the transaction service; the minimum connection service criteria establishes baseline requirements for a connection between each of the first device and the second device and the transaction service to be used for exchange of transaction data; and in response to the second device not having a second connection that meets the minimum connection service criteria, the second device utilizes the first device as an intermediary device to transmit the second device transaction data to the transaction service.

US Pat. No. 10,970,686

TESTING POWER REUSE SYSTEM, POWER REUSE CIRCUIT AND TESTING POWER REUSE METHOD

FSP TECHNOLOGY INC., Tao...

1. A testing power reuse system, comprising:at least one testing device;
at least one virtual currency calculating circuit; and
a power reuse circuit, coupled to the at least one testing device through at least one first interface circuit and coupled to the at least one virtual currency calculating circuit through at least one second interface circuit,
wherein the power reuse circuit is configured to receive power generated by performing a testing operation on the at least one testing device through the at least one first interface circuit and provide the power to the at least one virtual currency calculating circuit through the at least one second interface circuit, and
the at least one virtual currency calculating circuit is driven by the power to perform a virtual currency calculating operation.

US Pat. No. 10,970,685

ELECTRONIC FUNDS TRANSFERS BASED ON AUTOMATIC CRYPTOCURRENCY TRANSACTIONS

Capital One Services, LLC...

1. A method, comprising:receiving, by a device, authentication information from a user,
wherein the authentication information comprises an access token;
receiving, by the device, transaction information from a transaction terminal,
the transaction information comprising a transaction amount;
determining, by the device, an amount of a first type of cryptocurrency or a second type of cryptocurrency based on the transaction amount;
receiving, by the device and from the user, a first rule associated with the first type of cryptocurrency and a second rule associated with the second type of cryptocurrency,
receiving, by the device, a credential from the user;
logging, by the device, into a cryptocurrency server using the credential;
accessing, by the device, at the cryptocurrency server a cryptocurrency account of the user, using the access token,
wherein the cryptocurrency account is associated with the first type of cryptocurrency and the second type of cryptocurrency;
selecting, by the device, based on the cryptocurrency account, and using the first rule or the second rule, a portion of the first type of cryptocurrency or the second type of cryptocurrency; and
selling, by the device, the selected portion at the cryptocurrency server.

US Pat. No. 10,970,684

SYSTEMS AND METHODS FOR MAINTAINING DEPOSITS OF MATH-BASED CURRENCY

Wells Fargo Bank, N.A., ...

1. A method of receiving deposited math based currency (“MBC”) at a financial institution, the method comprising:storing, by an MBC transaction processor of a plurality of processors, a plurality of private and public key pairs in a pooled database;
receiving, by an account balance processor of the of the plurality of processors, a deposit request from a customer via a customer computing device, the deposit request including a customer private key for an amount of MBC;
communicating, by the account balance processor, the customer private key to the MBC transaction processor of the plurality of processors, wherein the MBC transaction processor is communicably coupled to the account balance processor;
creating, by the MBC transaction processor, a first private and public key pair;
transferring, by the MBC transaction processor, a first transaction in the amount of MBC to the created first private and public key pair from the customer by signing a transaction request with the customer private key;
associating, by an overlay ledger, the amount of MBC with an account of the customer of the overlay ledger;
tracking, by the overlay ledger, an association of an amount of MBC with each of a plurality of MBC accounts of a plurality of customers;
updating, by the account balance processor, the overlay ledger by the amount of MBC; and
in response to transferring the first transaction, storing, by the MBC transaction processor, the first private and public key pair in the pooled database.

US Pat. No. 10,970,683

AUTOMATICALLY DETECTING AND STORING DIGITAL DATA ITEMS ASSOCIATED WITH DIGITAL CALENDAR ITEMS

ATLASSIAN PTY LTD., Sydn...

1. A computer-implemented method comprising:obtaining a digital calendar item for an event, the digital calendar item comprising time data, location data that includes an identifier that specifies a location of a phone conference call, a video feed, or a chat group associated with the event, and participant data for each of a plurality of participants;
determining, using the time data, that the event has begun, and in response to determining that the event has begun, using the location data to access the location of the phone conference call, the video feed, or the chat group associated with the event and determining which of the plurality of participants are present for the event using participant data that is associated with participant accounts or participant digital computing devices;
determining, during the event, that one participant shared a digital data item with the other participants as part of the event, and in response to determining that the participant shared the digital data item, at least transiently storing a reference or link to the digital data item;
determining that the event has ended, and in response to determining that the event has ended, updating the digital calendar item by adding the reference or link to the digital data item;
identifying a task assigned to a particular participant of the plurality of participants;
updating the digital calendar item to add the task that is assigned to the particular participant.

US Pat. No. 10,970,682

SYSTEM AND METHOD FOR AGENT SCHEDULING USING MIXED INTEGER LINEAR PROGRAMMING

INCONTACT, INC., Salt La...

1. A computer implemented method for agent scheduling for a plurality of agents, where agents' schedules are partly defined using tour groups comprising rules and templates, the method comprising:a. acquiring, by a computer, from one or more computer systems scheduling environment information including a schedule period, intervals during the schedule period agents can work, tour groups wherein each tour group includes a minimum and maximum number of work days constraint during the schedule period for all agent schedules created using the tour group, shifts templates, minimum and maximum number of agents to schedule using each tour group, and schedule period;
b. acquiring, by the computer, from one or more computer systems required staffing levels for the intervals during the scheduling period;
c. acquiring, by the computer, from one or more computer systems an objective criterion and a direction for optimization as minimization or maximization of the objective function value;
d. formulating, by the computer, a Mixed Integer Linear Program model with decision variables, an objective function based on the objective criterion acquired and constraints to enforce the tour group and shift template rules, wherein the objective function value evaluated for an integer feasible solution with the objective function of the Mixed Integer Linear Program model provides a measure to compare a plurality of integer feasible solutions to determine a best solution, and the requirement for meeting required staffing levels in a solution of the Mixed Integer Linear Program model;
e. executing, by the computer, the steps of a Mixed Integer Linear Program and Search Heuristic algorithm to generate integer feasible solutions to the Mixed Integer Linear Program model until one or more stopping conditions are satisfied, wherein the Mixed Integer Linear Program and Search Heuristic algorithm repeats the steps of the Mixed Integer Linear Program solutions algorithm to find a plurality of solutions to the Mixed Integer Linear Program model and executes, for each said solution by the Mixed Integer Linear Program solution algorithm, the steps of the search heuristic to find an integer feasible solution to the Mixed Integer Linear Program model;
f. generating, by the computer, agent schedules based on the Mixed Integer Linear Program model found by the Mixed Integer Linear Program and Search Heuristic algorithm; and
g. inputting, by the computer, the agent schedules to one or more computer systems for rostering.

US Pat. No. 10,970,681

CAMERA GESTURE CLOCK IN

WALMART APOLLO, LLC, Ben...

1. An employee timekeeping system for keeping track of the work time of an employee, comprising:an image acquisition device adapted to acquire a plurality of two-dimensional (2D) full-body images of an employee performing a gesture having unique motions during a predetermined acquisition window;
an image analysis device adapted to receive the plurality of 2D full-body images and extract features in the acquired 2D full-body images;
a timekeeping device for providing a date and time when the 2D full-body images were acquired;
a gesture timing device coupled to the image acquisition device, adapted to indicate a start time and an end time of an image acquisition window in which the image acquisition device acquires the 2D full-body images of the employee gesture;
an image database having a plurality of unique prestored features of the plurality of full-body images created by and associated with each employee;
a time and attendance database for storing when each employee started and ended work; and
a controller coupled to the image acquisition device, the image analysis device, the image database and the time and attendance database that:
receives the plurality of 2D full-body images acquired from the image acquisition device,
receives a date and time from the timekeeping device when the 2D full-body images were acquired,
provides the plurality of acquired 2D full-body images to the image analysis device,
receives a set of features from image analysis device,
searches image database for similar sets of features,
determines if the features of the 2D full-body images of the employee gesture match prestored features of a gesture closer than a predetermined error,
identifies an employee associated with the matching prestored features, if there is a match, and
stores the identification of the employee and the date/time of the image acquisition in a time and attendance database, if there is a match;
wherein the controller causes the image analysis device to match features of a plurality of 2D full-body images of the employee gesture to features of a plurality of 2D full-body images previously stored in the image database.

US Pat. No. 10,970,680

SYSTEMS AND METHODS FOR MONITORED SOCIAL MEDIA PARTICIPATION

United Services Automobil...

1. A system comprising:a processor; and
memory comprising computer-readable instructions that, when executed by the processor, cause the processor to effectuate operations comprising:
monitoring social media stream using a social media portal (SMP) by receiving the social media stream and an output lead, wherein the social media portal is coupled to a database used for storing social media posts, wherein the SMP comprises at least a stream capture engine, a database access engine, a search engine, and a scoring engine;
determining that a post of the social media posts on a social media network is associated with a trigger event by determining a plurality of scores associated with the likelihood that a plurality of posts on the social media network are associated with the trigger event, wherein the score represents calculated similarity between search criteria and the social media posts, and wherein the trigger event comprises life events, complaint events and insurance events;
routing, the social media posts based on output of the scoring engine, wherein a highest score represents a greatest likelihood the trigger event is associated with a social media post;
generating a lead depending on which trigger event is calculated to have the highest score;
filtering, using classification, clustering and recommendation machine learning algorithms, based at least on the plurality of scores, the plurality of social media posts;
transmitting a request for a response to the filtered plurality of social media posts;
transmitting, based at least on the determining that the social media post is associated with the trigger event and to a computing device, the post;
receiving, from the computing device, an indication of a social media account;
adding the response to a data store comprising a plurality of responses to the plurality of posts; and
posting, based at least on the plurality of responses being approved, the plurality of responses to the social media network under the social media account.

US Pat. No. 10,970,679

PRESENTING PROJECT DATA MANAGED BY A CONTENT MANAGEMENT SYSTEM

Dropbox, Inc., San Franc...

1. A computer-implemented method comprising:storing, in a content management system, a first content item at a first location associated with a first user account;
monitoring, by the content management system, user behavior with respect to one or more of content items, projects, or user communications associated with the first user account;
storing, by the content management system, user behavior data describing the user behavior at the first location associated with the first user account;
receiving, by the content management system, a request to create a first project;
in response to receiving the request, generating the first project, the first project including the first user account as a member of the first project;
based on the request and the stored user behavior data, recommending, by the content management system, an affiliation between the first project and the first content item;
upon receiving an approval to create the affiliation between the first project and the first content item, adding the first content item as a project item for the first project by:
generating a reference to the first content item, the reference to the first content item providing access to the first content item at the first location, and
generating a project user interface, the project user interface comprising user interface data representing project items in the first project and members of the first project.

US Pat. No. 10,970,678

CONFERENCE INFORMATION ACCUMULATING APPARATUS, METHOD, AND COMPUTER PROGRAM PRODUCT

KABUSHIKI KAISHA TOSHIBA,...

1. A conference information accumulating apparatus for accumulating conference information, the apparatus comprising:a generator configured to generate a user interface screen either for creating minutes of a conference based on the conference information or for viewing the created minutes; and
a calculator configured to detect a piece of the conference information correlated with a written text that is text written to the user interface screen by a minutes creator based on a predetermined operation targeting on the piece of the conference information including at least one of a searching of the piece of the conference information, a replaying of the piece of the conference information, and a correcting of the piece of the conference information performed using the user interface screen by the minutes creator before the written text is written to the user interface screen, wherein
the generator generates, upon detection of the piece of the conference information that is correlated with the written text, the user interface screen enabling a reference to the piece of the conference information, the user interface screen including a first area for referring to the conference information and a second area for creating or viewing the minutes, the second area having a first operation button arranged therein with which the piece of the conference information is transformed and mapped to the written text that is summarized text of the piece of the conference information, wherein the piece of the conference information is transformed and mapped based on determining a user identity within uttered speech that is included in the conference information and an uttered action item that is uttered in near contemporaneity with the user identity, associating the uttered action item with the user identity, and transcribing, based on the uttered speech, the user identity and the uttered action item into the summarized text, and
the generator generates, in response to an operation of the first operation button, the user interface screen in which the first area displays the piece of the conference information that is correlated with the written text.

US Pat. No. 10,970,677

MANAGING UPDATES FROM REFERENCE LABORATORIES

CERNER INNOVATION, INC., ...

1. Non-transitory computer storage media having computer-executable instructions embodied thereon that, when executed by a server having a processor and memory communicating through a computer network, facilitate an improved method utilizing a database for storing information to facilitate a mapping process of providing a centralized conduit for leveraging a centralized mapping between organizations and managing updates from reference laboratories, the method comprising:receiving by the server, updates including changes to requirements for attributes of one or more tests from at least one reference laboratory, the updates comprising changes to the requirements for a collection of specimens for the one or more tests including at least one of a specimen container type, a collection volume, a collection method, or an acceptable results range, the updates including information unique to one of the at least one reference laboratory, wherein the attributes of the one or more tests include one or more of method, minimum volume, specimen collection, specimen container, specimen storage, specimen type, specimen volume, special instructions, and testing frequency, and wherein the at least one reference laboratory is a location for performing the one or more tests;
mapping by the server, the changes to the requirements for the attributes of the one or more tests received from the at least one reference laboratory to at least one client, the mapping comprising identifying the at least one client associated with the at least one reference laboratory based on a network identification associated with the at least one reference laboratory and the at least one client and determining that at least one of the updates including at least one of the changes to the requirements for the attributes of one or more tests received from the at least one reference laboratory is relevant to the at least one client based on the at least one client's prior use of the one or more tests for which the updates were received, the mapping further comprising identifying an alias associated with the test used by the reference laboratory or the client, wherein the alias is a code that provides a pointer to the test in a database used by the reference laboratory or the client;
based on the determined relevance to the at least one client, sending by the server the at least one of the updates, the alias, and a determination of a status of the at least one of the updates to the at least one client; and
based on the determined relevance to the at least one client, applying by the server the at least one of the updates by modifying the attributes of the one or more tests such that the information corresponding to the at least one of the updates is available for use by the at least one client when presented to the at least one client.

US Pat. No. 10,970,676

VEHICLE INVENTORY AND CUSTOMER RELATION MANAGEMENT SYSTEM AND METHOD

VOXX INTERNATIONAL CORPOR...

1. A method for managing a plurality of vehicles on a lot of a car dealer, the method comprising steps of:storing information in a database implemented on a database server of a first computer system, the information comprising staff user accounts of staff users of the car dealer, vehicle owner accounts of buyers of the vehicles, and vehicle accounts of the plurality of vehicles, the first computer system being accessible through a network;
operating a second computer system to provide secure access to the first computer system and to enable configuring within the database server the staff user accounts, the vehicle owner accounts, and the vehicle accounts;
configuring the second computer system and a plurality of handheld units, so that each handheld unit of the plurality of handheld units is configured for wireless communications with the second computer system, is capable of sending wirelessly to the second computer system one or more vehicle criteria inputted into said each handheld unit, and is capable of receiving wirelessly from the second computer system location of a vehicle of the plurality of vehicles matching the one or more vehicle criteria and one or more user access authorizations for accessing the vehicle matching the one or more vehicle criteria using said each handheld unit;
associating said each handheld unit with a unique staff user account of the staff user accounts; and
installing a plurality of modules in the plurality of vehicles, one module of the plurality of modules per vehicle of the plurality of vehicles, each module of the plurality of modules being in communication with a vehicle data bus system of the vehicle in which said each module is installed, said each module having a first operating mode for managing the vehicle in which said each module is installed on the lot before sale of the vehicle in which said each module is installed, and a second operating mode for communicating after sale by the car dealer of the vehicle in which said each module is installed;wherein:the staff user accounts comprise staff user identification data and a plurality of user access authorization levels for the plurality of vehicles, the access authorization levels providing permissions for vehicle door locks, ignition start, and feature configurations of the plurality of vehicles;
the vehicle owner accounts comprise vehicle owner identifications and transaction data associated with sales of the vehicles of the plurality of vehicles;
the vehicle accounts comprise vehicle identification data of the plurality of vehicles, the feature configurations of the plurality of vehicles, and vehicle location data of the plurality of vehicles;
when said each module is in the first operating mode,
said each module is configured to communicate wirelessly with the second computer system and to communicate wirelessly with the plurality of handheld units,
said each module is further configured to provide access to at least one item selected from the group consisting of door locks of the vehicle in which said each module is installed, ignition system of the vehicle in which said each module is installed, and device configuration system of the vehicle in which said each module is installed, if a proper authorization signal is received by said each module from a handheld unit of the plurality of handheld units matching an authorization configuration provided to said each module by the second computer system, and
said each module is further configured to communicate wirelessly vehicle location data to the second computer system; and
when said each module is placed in the second operating mode, the first operating mode is permanently disabled in said each module and said each module is configured to collect vehicle performance data from the vehicle data bus of the vehicle in which said each module is installed and to communicate through a cellphone the vehicle performance data to the first computer system.

US Pat. No. 10,970,675

DATA PROCESSING SYSTEMS FOR GENERATING AND POPULATING A DATA INVENTORY

OneTrust, LLC, Atlanta, ...

1. A non-transitory computer-readable medium for generating a data inventory for a plurality of inter-related data assets, the non-transitory computer-readable medium storing computer-executable instructions for:generating a data model mapping one or more relationships between a plurality of data assets, the data model comprising one or more data inventories, wherein each data inventory of the one or more data inventories is associated with a respective data asset of the plurality of data assets;
identifying a primary data asset from among the plurality of data assets that collects or stores personal data of one or more data subjects;
generating a data inventory for the primary data asset, the data inventory storing one or more primary data asset inventory attributes comprising transfer data associated with the primary data asset;
determining one or more individuals permitted to access the one or more pieces of the personal data of the one or more data subjects collected or stored by the primary data asset;
modifying the one or more primary data asset inventory attributes to include access data that identifies the one or more individuals permitted to access to the one or more pieces of the personal data of the one or more data subjects collected or stored by the primary data asset;
modifying the data model to include the data inventory for the primary data asset;
electronically linking the primary data asset to the data inventory for the primary data asset in the data model;
using the transfer data associated with the primary data asset to identify a transfer data asset from among the plurality of data assets;
at least partially in response to identifying the transfer data asset that has been identified using the transfer data associated with the primary data asset;
modifying the data inventory for the primary data asset to include both the transfer data asset and one or more transfer data asset inventory attributes comprising the transfer data associated with the primary data asset; and
modifying a data inventory for the transfer data asset to include the primary data asset;
digitally storing the modified data inventory for the primary data asset and the modified data inventory for the transfer data asset in computer memory; and
electronically linking the primary data asset and the transfer data asset in the data model.

US Pat. No. 10,970,674

MOBILE TABLET GUN SYSTEM WITH MOBILE TABLET HAVING A MOBILE POS AND ENTERPRISE RESOURCE PLANNING APPLICATION FOR POS CUSTOMER ORDER FULFILLMENT AND IN-STORE INVENTORY MANAGEMENT FOR RETAIL ESTABLISHMENT

RETAILTECHNOLOGIES CORPOR...

1. A mobile tablet gun system for interchangeably attaching and electronically coupling a mobile tablet device to a base form factor, wherein:a) said base form factor has a base mount universal receiver integrated with a rotational coupling unit connected to an interface printed circuit board (P.C.B.) located on said base form factor for interchangeably mounting and electronically communicating with said mobile tablet device and providing rotatable connection between said mobile tablet device and said base form factor, and said base mount universal receiver with rotational coupling is located between said mobile tablet device and said base form factor,
b) said base form factor has a bar code scanner and bar code scanner interface P.C.B. in communication therewith, and said base form factor includes a trigger adapted to activate said bar code scanner;
c) said mobile tablet device has a top housing with a display and a bottom housing, mobile application software and a Primary Printed Circuit Board (P.C.B.) for processing commands for running real-time custom retail applications, system applications and firmware executable on a Microsoft Windows operating system (O.S.) adapted to be connected to said base form factor through said base mount universal receiver and capable of receiving substantially the same remote Windows Updates as an industry-standard tablet, laptop or desktop point of sale (P.O.S.) terminal;
d) said mobile tablet device has one or more docking station P.C.B. interface connectors located on said bottom housing of said mobile tablet device for mounting and electronically connecting said mobile tablet device to said base form factor through said base mount universal receiver with said rotational coupling unit;
e) aid mobile tablet device has a battery assembly having one or more batteries;
f) said mobile tablet device has a universal serial bus (USB) payment card reader supporting a Magnetic Strip Reader (M.S.R.) and a Europay MasterCard Visa (E.M.V.) payment card reader for executing point of sale (P.O.S.) transactions and real-time daily chain store-level inventory management functions for such tasks as physical inventory, cycle counting, inventory receiving, store to store transfers, return to vendor, and product re-ticketing;
g) said mobile tablet device has a Near Field Communications (N.F.C.) USB module for executing “tap and go” contactless point of sale (P.O.S.) transactions and real-time daily chain store-level inventory management functions for such tasks as physical inventory, cycle counting, inventory receiving, store to store transfers, return to vendor, and product re-ticketing; and
h) said mobile tablet gun system instantly and seamlessly migrates between a variety of form factors, including a fixed P.O.S. terminal, a Wi-Fi handheld terminal and a cellular handheld terminal, enabling sales associates to service their customers without disruption of work flow or disruption of a P.O.S. transaction.

US Pat. No. 10,970,673

BILL OF MATERIAL SYNCHRONIZATION

INTERNATIONAL BUSINESS MA...

9. A method performed by a synchronization system incorporated in a physical computing system programmed to synchronize a source bill of material stored on a first networked physical computing system with a target bill of material stored on a second, separate networked physical computing system, the method comprising;with the physical computing system, monitor said source bill of material at said first networked physical computing system for revisions;
in response to a determination by said physical computing system that multiple revisions have been applied to a source node from a tree structure representing said source bill of material at said first networked physical computing system after a corresponding target node from a tree structure representing said target bill of material at said second networked physical computing system was last updated, creating a copy of said target node and revising said copy of said target node with revisions based on a revision history associated with said source bill of material; and
synchronizing the two bills of material by updating, with said physical computing system, said target node at said second networked physical computing system with said revisions only when the revised copy of said target node matches said source node,
so that manufacturing that is based on said target bill of materials located at said second networked physical computing system is accurate with respect to said source bill of materials located at said first networked physical computing system.

US Pat. No. 10,970,672

INVENTORY MANAGEMENT SYSTEM AND METHOD

Alert Innovation Inc., N...

1. A system for implementing a fully automated secure supply chain for end to end tracking of eaches, comprising:one or more distribution centers for receiving eaches from a manufacturer and delivering the eaches to a market;
a plurality of robots at the one or more distribution centers for sorting eaches for shipment between the one or more distribution centers and the market, the plurality of robots comprising:
a first group of one or more robots configured to accept pallets comprising cases of caches;
a second group of one or more robots configured to selectively retrieve stored caches to fill a market order;
a plurality of tracking sensors included on the first and second groups of robots for capturing identification information regarding the eaches; and
a tracking system, comprising a unified commerce engine, for receiving and storing the identification information from the tracking sensors, the tracking system tracking locations of eaches from delivery from the manufacturer to the one or more distribution centers and from the one or more distribution centers to delivery to the market using the identification information.

US Pat. No. 10,970,671

WIRELESS AUTHENTICATION METHOD AND DEVICE OF RECYCLABLE LOGISTICS APPARATUS

SHANGHAI HONGYAN RETURNAB...

1. A wireless authentication method of recyclable logistics apparatus, the recyclable logistics apparatus comprises a wireless beacon unit which periodically transmits a broadcast frame, the broadcast frame at least includes a unique identity of the wireless beacon unit,a plurality of recyclable logistics apparatuses is capable of being stacked as a plurality of stacks; when stacked, wireless beacon units built in respective recyclable logistics apparatuses are connected with contact points between recyclable logistics apparatuses components and between the recyclable logistics apparatuses through wires, thereby forming a parallel structure of circuits; and the broadcast frame further includes identification information of the wireless beacon unit, if the identification information of the wireless beacon unit of one recyclable logistics apparatus in a stack is changed, the identification information of the wireless beacon units of all the recyclable logistics apparatuses in the stack are all changed;
wherein, the wireless authentication method comprises the following steps:
first mobile terminal receives a first broadcast frame and uploads information collected from the first broadcast frame as first upload information to a cloud server, the cloud server generates authentication information according to an authentication rule;
the cloud server associates the first upload information with the authentication information and stores them in a database; and
the cloud server transmits the authentication information to second mobile terminal, the second mobile terminal receives a second broadcast frame, and through the authentication information, information collected from the second broadcast frame is compared with the information collected from the first broadcast frame which is associated with the authentication information;
wherein in the step that through the authentication information, information collected from the second broadcast frame is compared with the information collected from the first broadcast frame which is associated with the authentication information, the second mobile terminal also uploads the information collected from the second broadcast frame to the cloud server so that the cloud server performs the comparison; or
wherein in the step that through the authentication information, information collected from the second broadcast frame is compared with the information collected from the first broadcast frame which is associated with the authentication information, the second mobile terminal obtains the information collected from the first broadcast frame which is associated with the authentication information from the cloud server so that the second mobile terminal performs the comparison.

US Pat. No. 10,970,670

YARD MANAGEMENT SYSTEM

Walmart Apollo, LLC, Ben...

1. A yard management system for a distribution center, the system comprising:a database holding load information related to a plurality of loads stored in a cargo trailer destined for the distribution center, the load information including an identification associated with each load in the plurality of loads, and additional load information related to a driver delivering the cargo trailer to the distribution center, a vehicle transporting the cargo trailer, a freight-type associated with each of the plurality of loads, and estimated time of arrival;
a mobile application associated with a user account and executable on a mobile device, the mobile application when executed:
acquires at least one of the plurality of identifications of the load;
queries the database, with the acquired at least one of the plurality of identifications of the load, for the load information related to the load;
receives the queried load information from the database;
updates the additional load information based on information of load documentation; and
generates a machine-readable representation corresponding to the at least one of the plurality of identifications of the load;
an autonomous yard vehicle including a first coupling configured to mechanical couple with a second coupling of a cargo trailer which includes a plurality of loads;
a computing system in communication with the mobile device, the database, and the autonomous yard vehicle, the computing system when executed:
receives the updated additional load information of the plurality of loads from the mobile device;
updates the estimated time of arrival (ETA) of the cargo trailer based on the updated additional load information received from the mobile device;
transmits the updated ETA to the mobile device associated with the user account;
receives, from a client device at a gate of the distribution center, an image of the machine-readable representation scanned from the mobile device by the client device upon arrival of the cargo trailer at the gate;
extracts the at least one of the plurality of identifications of the load from the scanned image;
renders a map of the geographic area including a transfer location at which the cargo trailer is to be dropped off after passing through the gate; and
transmits the map and a load status to the mobile device associated with the user account and the autonomous yard vehicle;
wherein, in response to detecting the mobile device at the transfer location, the autonomous yard vehicle navigates to the transfer location based on the map to facilitate mechanical coupling of the autonomous yard vehicle and the cargo trailer and to drive the autonomous yard vehicle to one of a holding location or a dock location based on the plurality of loads in the cargo trailer.

US Pat. No. 10,970,669

BLOCKCHAIN ENABLED TRANSACTION PROCESSING FOR AN INDUSTRIAL ASSET SUPPLY CHAIN

GENERAL ELECTRIC COMPANY,...

1. A system to facilitate transaction processing associated with an industrial asset supply chain having a first entity and a second entity, comprising:a first entity database storing electronic records including information associated with pre-delivery data about an industrial asset;
a first entity communication port to exchange information via a distributed computer system;
a first entity computer processor coupled to the first entity database and the first entity communication port and adapted to:
retrieve from the first entity database the information associated with the pre-delivery data about the industrial asset, and
record pre-delivery data about the industrial asset via a secure, distributed transaction ledger;
a second entity database storing electronic records including information associated with a post-delivery event involving the industrial asset;
a second entity communication port to exchange information via the distributed computer system; and
a second entity computer processor coupled to the second entity database and the second entity communication port and adapted to:
retrieve from the second entity database the information associated with the post-delivery event involving the industrial asset,
record post-delivery event data about the industrial asset via the secure, distributed transaction ledger; and
a transaction prediction platform associated with the secure, distributed transaction ledger and adapted to trigger one or more payments for the industrial asset supply chain based on the pre-delivery data and the post-delivery event data.

US Pat. No. 10,970,668

DEPOT DISPATCH PROTOCOL FOR AUTONOMOUS LAST-MILE DELIVERIES

DoorDash, Inc., San Fran...

1. A method comprising:receiving a first order message from a first customer device corresponding to a first customer, the first order message including a first timestamp and a first order for real-time delivery of perishable goods from a first merchant, wherein the first customer is located in a delivery area;
training a neural network to dynamically generate estimated time of arrivals (ETAs) for one or more first events for the first order, including a first ETA, wherein the neural network comprises a plurality of computational layers including weighted parameters, wherein training the neural network includes:
receiving a dataset including a series of successive training events with corresponding known time durations between training events, and
inputting different combinations of training events and time durations, and comparing generated ETA predictions with known time durations between training events to automatically and iteratively adjust the weighted parameters in the neural network;
inputting the first timestamp corresponding to the first order message into the neural network to automatically generate the first ETA using the weighted parameters;
transmitting a first order notification to a first courier device corresponding to a first merchant courier based on the first ETA, wherein the first order notification includes: a location of the first merchant and a location of a merchant depot, wherein the first order is grouped into a first batch of orders at the merchant depot, wherein the first batch of orders includes one or more orders and is grouped based upon a proximity of delivery destinations of the one or more orders in the first batch of orders to a customer depot; and
transmitting a routing notification to a second courier device corresponding to a depot courier based on one or more of the ETAs, wherein the routing notification includes a location of the customer depot corresponding to the delivery area.

US Pat. No. 10,970,667

SYSTEM AND METHOD FOR DYNAMICALLY ROUTING ITEMS

UNITED STATES POSTAL SERV...

1. A system for handling items comprising:an inventory control processor configured to receive and store item information including a destination for each of a plurality of items and configured to generate a run manifest including the item information;
a memory configured to store a plurality of static routes, each of the static routes including a set of fixed delivery destinations to be followed in a fixed order, the plurality of static routes including all delivery destinations within a specific geographic location;
a routing processor configured to:
receive the run manifest;
group delivery destinations from the plurality of static routes into a plurality of subsets according to a quantity of items to be delivered to one or more of the delivery destinations from the plurality of static routes; and
generate a plurality of dynamic routes, each of the plurality of the dynamic routes based upon the plurality of subsets;
assign each of the quantity of items to one of the plurality of dynamic routes; and
generate a sequence identifier for each of the quantity of items according to the assigned one of the plurality of dynamic routes;
a processor configured to associate an item identifier for each of the quantity of items with the plurality of dynamic routes to which the quantity of items have been assigned and to store the association in a memory; and
a sorter comprising a scanning device, the sorter configured to receive the quantity of items and the association and to sort the quantity of items according to the assigned dynamic routes.

US Pat. No. 10,970,666

TRANSPORTATION AND SHIPPING MANAGEMENT SYSTEM

Askul Corporation, Tokyo...

1. A transportation and shipping management system for managing a selling system for shipping out an item from a warehouse or a nearest base associated with a delivery destination in accordance with the delivery destination,the transportation and shipping management system comprising an instruction unit configured:
to, before receiving an order placement for an item from a customer, transmit, to a warehouse server associated with the warehouse or a relay base server associated with a relay base, a transportation instruction for transporting the item to the nearest base from the warehouse or the relay base, and
to, after receiving the order placement for the item from the customer, transmit, to a nearest base server associated with the nearest base, a submission instruction for submitting the ordered item to the customer;
wherein the item is present in plurality and at least a part of the items is stored in plurality in a first container;
wherein the submission instruction includes an instruction for carrying the first container, which is temporarily stored in the nearest base and in which the plurality of items are stored, to the delivery destination and collecting a second container located at the delivery destination; and
wherein the transportation and shipping management system further comprises a server comprising a billing processing unit configured to automatically generate billing information on the basis of a difference between a reading value received from a weight sensor of the first container transmitted via a communication device of the first container, and a reading value received from a weight sensor of the second container transmitted via a communication device of the second container.

US Pat. No. 10,970,665

LOGISTICS INFORMATION ACQUISITION METHOD AND SYSTEM FOR TRANSNATIONAL TRANSPORT

DEMON NETWORK TECH. CO., ...

1. A method for acquiring logistics information of transnational transport, characterized in that the method comprises:step A, receiving a tracking number inputted by a user through a website query portal, and recognizing information and an identifier of a starting transporter in the tracking number based on a preset tracking number rule base;
step B, acquiring logistics information of the starting transporter based on the tracking number, the information and the identifier of the starting transporter;
step C, extracting a logistics information query parameter of a transporter in a first transit country, information and an identifier of the transporter in the first transit country from the logistics information of the starting transporter;
step D, acquiring logistics information of the transporter in the first transit country based on the logistics information query parameter of the transporter in the first transit country, the information and the identifier of the transporter in the first transit country;
step E, extracting a logistics information query parameter of a transporter in a second transit country, information and an identifier of the transporter in the second transit country from the logistics information of the transporter in the first transit country;
step F, acquiring logistics information of the transporter in the second transit country based on the logistics information query parameter of the transporter in the second transit country, the information and the identifier of the transporter in the second transit country;
step G, repeating step E and step F until no logistics information query parameter of a next country is contained in logistics information of a transporter in a transit country, and taking logistics information of the transporter in the transit country as logistics information of a final transporter;
step H, combining and outputting the logistics information of the starting transporter, the logistics information of the transporters in all the transit countries and the logistics information of the final transporter and the identifiers of all the transporters;
classifying the combined and outputted logistics information into different types, different languages, different formats or different description of the logistics information, to obtain states of different stages in a logistics process; and
unifying the classified states into a preset data format, and outputting and displaying the preset data format of states;
wherein the acquiring the logistics information of the transporter in the step B, the step D or the step F comprises:
searching a preset rule base for an information acquisition manner corresponding to the transporter based on the information of the transporter, wherein the information acquisition manner comprises: a simulative request on a website of the transporter, a request at a public query interface of the transporter or a request at a cooperative query interface of the transporter;
acquiring data related to the logistics information query parameter based on the information acquisition manner through network access; and
extracting all information related to logistics information from the data upon acquiring the data, wherein the logistics information comprises a logistics event, an event time, an event location, various significant identifiers or information of a next transit country.

US Pat. No. 10,970,664

IMPAIRMENT DETECTION SYSTEM AND METHOD

Predictive Safety SRP, In...

1. A computer-implemented method, executed on a computing device, comprising:administering an alertness test on a client electronic device to determine a result for a user, wherein the result of the alertness test is indicative of a level of alertness of the user, wherein administering the alertness test to the user includes:
rendering a plurality of objects for use within the alertness test being administered to the user;
rendering a disrupter configured to distract the user, wherein the disrupter is configured to rotate over the plurality of objects and, while rotating, temporarily obscure at least a portion of one or more objects of the plurality of objects during the alertness test;
soliciting a response from the user concerning whether any object of the plurality of objects is different from the other objects;
receiving the response from the user; and
determining the result of the alertness test for the user based, at least in part, upon the response received from the user;
receiving biometric information concerning the user from a biometric device; and
adjusting the result of the alertness test for the user based, at least in part, upon the biometric information, wherein adjusting the result of the alertness test includes one or more of adjusting the result of the alertness test upwards indicative of a greater level of alertness of the user and adjusting the result of the alertness test downwards indicative of a lower level of alertness of the user.

US Pat. No. 10,970,663

SEGMENTED SOCIAL COLLABORATION WITH COMMUNICATION ARBITRATION

International Business Ma...

1. A computer-implemented method comprising:determining a workflow segmentation model for a user, wherein the workflow segmentation model includes a plurality of task contexts;
receiving, from a sender client device, one or more communication events destined for a receiver client device associated with the user;
performing a computerized natural language analysis process to identify keywords from each of the one or more communication events based on the plurality of task contexts; and
arbitrating, based on the workflow segmentation model, the one or more communication events sent to the user, wherein arbitrating includes determining a correlation value, based on a number of matching criteria of the communication event in a task context table;
determining a communication event as a mismatched communication event in response to the correlation value being below a predetermined threshold; and
in response to identifying a mismatched communication event:
deferring delivery of the communication event based on an induced communication delay;
generating an auto-reply message, wherein the auto-reply message includes a future delivery time of the mismatched communication event, wherein the future delivery time is based on the induced communication delay; and
sending the auto-reply message to the sender client device and
in response to data from an environmental sensor exceeding a predetermined value, overriding the induced communication delay, and delivering the mismatched communication event to the receiver client device immediately; and
in response to data from the environmental sensor not exceeding a predetermined value, delivering the mismatched communication event to the receiver client device upon expiration of the induced communication delay.

US Pat. No. 10,970,662

METHOD AND SYSTEM FOR LOCATING RESOURCES AND COMMUNICATING WITHIN AN ENTERPRISE

Steelcase Inc., Grand Ra...

1. A method of generating views for use in a resource management system in an entity comprising one or more facility and a plurality of data collectors throughout the entity, the resource management system comprising a database of a plurality of entity resources registered with the resource management system including an identifier correspond to each resource, and a database identifying a location of each of the plurality of data collectors, the method comprising the following steps:associating tracking devices with each of the plurality of entity resources including a resource identifier;
sensing characteristics of the resource with the corresponding tracking device;
transmitting the resource identifier and sensed characteristics from the tracking devices to the resource management system;
transmitting the identifier to one or more of the plurality of data collectors in a vicinity of the transmitting device;
transmitting data from the data collectors to the resource management system;
using the data to generate views of the resources.

US Pat. No. 10,970,661

SYSTEM AND METHOD FOR MONITORING MOTION AND ORIENTATION PATTERNS ASSOCIATED TO PHYSICAL ACTIVITIES OF USERS

Racefit International Com...

1. A system for monitoring motion and orientation patterns associated to one or more physical activities of a user, the system comprising:one or more sensors, wherein the one or more sensors are positioned within an upper garment and a lower garment of the user;
a processor in communication with the one or more sensors;
a memory coupled with the processor, wherein the processor executes programmed instructions stored in the memory for:
capturing data associated to one or more physical activities executed by a user from the one or more sensors, wherein the physical activities comprise a static posture and one or more dynamic movements, wherein the data comprises at least motion and orientation data, wherein the motion and the orientation data include a tilting angle, a curvature angle, an extension angle, and a flexion angle;
comparing the data with a plurality of pre-defined motion patterns in order to identify a matched motion pattern, wherein each pre-defined motion pattern comprises a first reference threshold value and a second reference threshold value, wherein the first reference threshold value and the second reference threshold value indicates a baseline level and a minimum passing level, respectively, required for a successful execution of the one or more physical activities, wherein the baseline level and the minimum passing level is determined based upon static and dynamic calibration techniques respectively, wherein the static calibration technique is used when the user is instructed to perform the static posture and the dynamic calibration technique is used when the user is instructed to perform the dynamic movements, wherein the static and/or dynamic calibration techniques enable in generating a profile of baseline levels and minimum passing levels customized for different users depending on posture, size and physical ability of the different users;
determining, based upon the data, a first threshold value and a second threshold value pertaining to the one or more physical activities executed by the user;
comparing the first threshold value and the second threshold value with the first reference threshold value and the second reference threshold value, respectively, corresponding to the matched motion pattern; and
notifying an output to the user based upon the comparison of the first threshold value and the second threshold value with the first reference threshold value and the second reference threshold value, wherein the output is indicative of a performance of the user in executing the one or more physical activities;
wherein the data captured from the one or more sensors is converted into a fuzzified data using a knowledge base.

US Pat. No. 10,970,660

COMMUNITY-BASED DATA ANALYSIS IN A SOFTWARE-AS-A-SERVICE SYSTEM USING A DEEP LEARNING CLASSIFIER

Coupa Software Incorporat...

1. A computer implemented method executed using two or more spend management cloud systems (SMCS) each in a different region and coupled to regional data/object storage and communicatively coupled to computers of suppliers and computers of customers and to a plurality of enterprise resource planning (ERP) systems via a data network, the customers comprising a particular customer and a plurality of customers, the method comprising:importing spend data from the ERP systems into the SMCS via a REST API, the spend data comprising a plurality of spend data lines from both the computer of the particular customer and the computers of the plurality of customers;
using a trained deep learning classifier to classify the spend data lines in a set of standard spend categories, the trained deep learning classifier being programmed to perform sequence classification using a trained long short-term memory recurrent neural network (LSTM/RNN) being programmed using word vector representations and embedding layers to predict a category for an input sequence, the set of standard spend categories corresponding to set of standard commodity codes of a standard taxonomy of commodities, the spend data lines in part comprising inconsistent or new text descriptions of commodities in a particular spend category, the LSTM/RNN having been trained based on actual or synthetically generated spend data lines with corresponding spend category labels;
computing a first aggregate of monetary amounts from a first set of classified spend data lines associated with the particular customer having a user account with the spend management cloud service;
computing a second aggregate of monetary amounts from a second set of classified spend data lines associated with a community of the plurality of customers having separate user accounts with the spend management cloud service, the second aggregate of monetary amounts comprising community-based spend aggregates that are computed from a set of spend data that has been collected from the community comprising a plurality of regional SMCS instances, the second aggregate comprising one or more of: number of distinct individual commodities per family spend category per supplier, number of distinct customers per family spend category per supplier, average per-customer spend amount, average per-customer transaction amount, total spend amount, total spend transaction amount, each being associated with one of the spend categories;
causing display of a graphical user interface that includes a visible representation of the first aggregate monetary amount for comparison with a visible representation of the second monetary amount; and
wherein the method is performed by a computing system having one or more processors and storage media storing one or more programs, the one or more programs including instructions configured to perform the method.

US Pat. No. 10,970,659

TASKING SYSTEM

Hipaax L.L.C., Mason, OH...

1. A tasking system that uses various communication modems to allow management to interact with one or more registered individuals and to assign, track and monitor tasks assigned to a qualified registered individual, the tasking system comprising:an administrative control system having individual information and tasking information for performing a task;
one or more communication apparatus in communication with said administrative control system and operates to receive tasking information from said administrative control system and transmitting response information to said administrative control system;
a communication network for providing communication between said administrative control system and said one or more individual communication apparatus;
a tracking system that operates to track the location of one or more registered individuals within a defined area and transmits tracking information to said administrative control system;
wherein said administrative control system operates to identify qualified registered individuals based on said individual information, said tasking information and said tracking information and transmits a task and tasking information to said one or more qualified registered individuals;
wherein at least one communication apparatus operates to receive said task and said tasking information from said administrative control system and transmits an acceptance and any response information from one said qualified registered individual who accepted said task to said administrative control system;
at least one said one or more portable communication apparatus includes one or more biological sensors that operates to obtain biological information of said qualified registered individual who accepted said task when performing said task and transmits said biological information to said administrative control system whereby said administrative control system then operates to compare said biological information to a biological threshold reference that indicates when an individual is approaching or exceeding a physical condition that may be harmful and if said biological information exceeds said biological threshold, said administrative control system transmits a warning notice to an output device to inform an administrator and to said qualified registered individual who accepted said task and is performing said task and a notice to other said registered individuals within said defined area to lend assistance in performing said task; and
wherein when said administrative control system receives said acceptance of said task it operates to remove said tasking information from individual communication apparatus of other said identified qualified registered individuals.

US Pat. No. 10,970,658

TECHNIQUES FOR BEHAVIORAL PAIRING IN A DISPATCH CENTER SYSTEM

Afiniti, Ltd., Hamilton ...

1. A method for behavioral pairing in a dispatch center system comprising:processing, by at least one dispatch center communicatively coupled to and configured to support behavioral pairing operations in the dispatch center system, a dispatch request for a customer based upon a customer request received at the dispatch center, wherein the dispatch center is configured to route the dispatch request to at least one queueing or switching component in the dispatch center system;
determining, by the at least one dispatch center, a plurality of field agents available to service the customer's dispatch request based upon the plurality of field agents assigned to the dispatch center;
generating, by at least one dispatch behavioral pairing module communicatively coupled to and configured to perform behavioral pairing operations in the dispatch center system, a model of preferred field agent-customer pairings based at least in part on historical field agent-customer interaction outcome data, wherein the model is determined further based upon information regarding the dispatch request and the plurality of field agents that is received from the dispatch center at the dispatch behavioral pairing module, wherein the model is used to increase the overall performance of the dispatch center system by causing the at least one queueing or switching component to more efficiently operate while targeting a balanced utilization of field agents;
selecting, by the at least one dispatch behavioral pairing module, one of the plurality of field agents based on the model; and
outputting, by the at least one dispatch behavioral pairing module, the selection to facilitate routing the dispatch request to the at least one queueing or switching component in the dispatch center system and dispatching the selected field agent to the customer.

US Pat. No. 10,970,657

METHOD AND SYSTEM FOR CAUSAL ANALYSIS OF OPERATIONAL OUTCOMES

Hublsoft Group Limited, ...

1. A non-transitory system for performing the causal analysis of at least one outcome of an organisation tangibly embodying an operating model and a causal model, wherein said system comprises:data processing means for defining at least one said outcome of the causal analysis;
a user interface in the form of a display screen by which the user may interact by placing their hand on or over a specific property or outcome displayed on the display screen, thereby defining a path to the said at least one outcome, said path comprising a number of links and a plurality of selected properties of the organisation operations, said properties selected from a larger number of available properties as a result of the said user interaction with said user interface and the data processing performed in the operating model as a result of the user interaction said available properties located within the operating model and representing the properties of interest in at least part of the said organisation, and identified in the operating model as believed to influence the said at least one outcome using organisation data, knowledge within the organisation and knowledge external to the organisation;
a data storage facility connected to the said operating model to access data showing current and previous states of factors of the organisation and selecting to use the same as inputs to the causal model calculations used across the path of selected properties for predicting the performance of the user selected outcome and wherein the control means for the display screen generates a first visual representation of the outcome and path including said links and selected properties in response to the first user interaction with the display screen and generates further visual representations relating to the same outcome and the configuration of the said further visual representations is determined by user interaction with the display screen to select one or more of the properties displayed on the first or further visual representation displayed thereon at that time and;
said causal model includes a time function to allow an outcome to be generated which is linked to the time function and data allocated on a time basis in the said database so as to generate on the display screen an indication of the current state of the outcome and a previous state of the outcome and wherein the said organisation operating model context is adjusted through filtering and selection to show outcomes related to different parts of the business operations and interventions are selectively made to make a structural change to the causal model through a prospective change to the business operating model, and hence to the properties being used in the causal model.

US Pat. No. 10,970,656

AUTOMATICALLY SUGGESTING PROJECT AFFILIATIONS

Dropbox, Inc., San Franc...

1. A method comprising:receiving, by a content management system, an identification of a project and a potential project item;
computing a relevance score that estimates a level of relevance between the potential project item and the project, wherein the potential project item is a content item and wherein the content management system generates the relevance score by performing one or more of:
a comparison of words in the content item and a description or title of the project;
a comparison of words in the content item and words in other content items associated with the project;
a comparison of users who have accessed the content item and members of the project; or
a comparison of users who have had the content item shared with them and members of the project,
wherein each comparison measures a level of similarity between the project and the potential project item;
determining that the relevance score is above a threshold; and
in response to the determining that the relevance score is above the threshold, providing a suggestion to create an association between the potential project item and the project.

US Pat. No. 10,970,655

SURGERY PLANNING BASED ON PREDICTED RESULTS

1. A method for planning a current medical procedure to be performed on a body part of a current patient, the method comprising:obtaining, using a computer, current first representation of a surface of the body part of the current patient;
retrieving, using the computer, at least one record from a database of records of previously performed medical procedures of previous patients, in which each record includes parameters of a previously performed procedure of said previously performed medical procedures, a first representation of the body part of the previous patient prior to performance of the previously performed procedure, and a second representation of a body part of the previous patient after the performance of the previously performed procedure, said at least one record being a best matching record to the current first representation of the current patient based on a similarity criterion between said current first representation of the first patient and said first representation of the body part of said previous patient;
creating a deformation map to characterize the transformation between the first and second representations of the body part of the previous patient, said map representing displacement between each of a plurality of points on the first representation of the previous patient and corresponding points on the second representation of the previous patient; and
applying the deformation map to the first representation of the surface of the body part of the current patient to predict the transformation for the current patient.

US Pat. No. 10,970,654

APPROPRIATELY DISTRIBUTING ITEMS USING AUGMENTED REALITY

International Business Ma...

1. A method for placing items in a container, the method comprising:receiving, by a point of sale terminal, a list of detected items that need to be placed in one or more containers, wherein said point of sale terminal interfaces with payment cards to make electronic funds transfers, wherein said list of detected items comprises items placed in a cart to be purchased by a customer that are detected via a radio-frequency identification (RFID) reader, wherein said RFID reader scans RFID tags on said items placed in said cart, wherein each of said RFID tags comprises an item identifier used to identify the item, wherein said list of detected items is maintained in a control module of said cart, wherein said list of detected items is transmitted from said control module to a server via a network, wherein said point of sale terminal accesses said list of detected items from said server via said network;
obtaining, by said point of sale terminal, health information and/or present physical condition information of said customer from a smart wearable worn by said customer, wherein said smart wearable is configured to monitor a heart rate, a body temperature and a level of sweat of said customer, wherein said smart wearable transmits said health information and/or present physical condition information of said customer to said point of sale terminal via said network;
determining, by said point of sale terminal, dimensions and a maximum amount of weight of items each of said one or more containers can hold, wherein a type of said one or more containers that are used to hold said detected items is based on input received by said point of sale terminal from a cashier or a bagger, wherein information as to said dimensions and said maximum amount of weight of items said type of said one or more containers can hold is stored in a database connected to said point of sale terminal via said network, wherein said point of sale terminal acquires said information as to said dimensions and said maximum amount of weight of items said type of said one or more containers can hold from said database in order to ensure that said detected items are placed in a fewest number of containers;
determining, by said point of sale terminal, dimensions of each item of said detected items, wherein said database is pre-populated with dimensions of items to be purchased in a store, wherein said point of sale terminal obtains said dimensions of each item of said detected items which are identified by said item identifiers from said database, wherein said dimensions of each item of said detected items are used to ensure that a number of items that are placed in a first container does not exceed a dimension of said first container using said determined dimensions of said one or more containers so as to ensure that said detected items are placed in a fewest number of containers;
determining, by said point of sale terminal, an estimated weight of each item of said detected items, wherein said estimated weight of each item of said detected items is obtained via a weight sensor located on said cart, wherein said weight sensor comprises a weighing scale;
determining, by said point of sale terminal, a maximum amount of weight for said first container said customer can carry based on said obtained health information and/or present physical condition of said customer, wherein a condition is identified from said health information and/or present physical condition information of said customer provided by said smart wearable worn by said customer, wherein in response to identifying said condition, a percentage of a maximum weight said first container can hold corresponds to said determined maximum amount of weight for said first container said customer can carry;
applying, by said point of sale terminal, rules related to food safety to determine which detected items should not be placed in vicinity with other detected items, wherein said rules are pre-populated in said database accessible by said point of sale terminal, wherein said rules ensure that certain items are not placed in said first container next to one another in order to ensure food safety; and
displaying, by said point of sale terminal, a visual representation of a placement of said detected items in said one or more containers on an augmented reality device so that said detected items are placed in said one or more containers in a fewest number of containers;
wherein a robot arranges said detected items in said one or more containers according to said visual representation of said placement of said detected items.

US Pat. No. 10,970,653

CACHING AND UPDATING DATA RECORDS BASED ON RELATIONSHIPS BETWEEN THE DATA RECORDS

PRICELINE.COM LLC, Norwa...

1. A method comprising:sending, from a travel listing aggregation computer, a first request to one or more data source computers for a plurality of listings that describe a plurality of travel items;
wherein each travel listing describes a respective travel item;
receiving, at the travel listing aggregation computer, the plurality of listings that describe the plurality of travel items from the one or more data source computers;
storing the plurality of listings in a local data cache of the travel listing aggregation computer;
using digitally programmed logic of the travel listing aggregation computer, determining that a particular listing of the plurality of listings stored in the local data cache contains invalid data;
sending, from the travel listing aggregation computer, a second request to a particular data source computer of the one or more data source computers for updated data for the particular listing;
using digitally programmed logic of the travel listing aggregation computer, identifying one or more related listings that are associated with the particular listing;
using digitally programmed logic of the travel listing aggregation computer, sending one or more third requests to one or more target data source computers of the one or more data source computers for updated data for the one or more related listings.

US Pat. No. 10,970,652

SYSTEM AND METHOD FOR SELECTING A CANDIDATE TRANSFER APPARATUS

Hammel Companies, Inc., ...

1. A method for determining a transfer apparatus based on user preferences and at least a transfer apparatus archive using machine learning, the method comprising:receiving, by a computing device, at least a transfer invocation and a plurality of user preferences;
generating, by the computing device and for each candidate transfer apparatus of a plurality of candidate transfer apparatuses, a plurality of performance prognoses corresponding to the plurality of user preferences, wherein generating each plurality of performance prognoses further comprises:
receiving, for a corresponding candidate transfer apparatus of the plurality of candidate transfer apparatus, a candidate transfer apparatus archive;
training, as a function of the candidate transfer apparatus performance archive and a supervised machine-learning process, a candidate transfer apparatus model; and
generating the plurality of candidate transfer apparatus performance prognoses as a function of the candidate transfer apparatus model and the at least a transfer invocation;
selecting, by the computing device, a candidate transfer apparatus as a function of the user preferences, wherein selecting further comprises:
generating an objective function of the user preferences, wherein;
the objective function is a function with a solution set including the plurality of performance prognoses corresponding to each candidate transfer apparatus;
the objective function generates an output ranking performance prognoses; and
selecting a candidate transfer apparatus of the plurality of candidate transfer apparatuses for which the output maximizes the rank; and
providing, by the computing device, the selected candidate transfer apparatus to a user.

US Pat. No. 10,970,651

ANALYTIC SYSTEM FOR TWO-STAGE INTERACTIVE GRAPHICAL MODEL SELECTION

SAS Institute Inc., Cary...

1. A non-transitory computer-readable medium having stored thereon computer-readable instructions that when executed by a computing device cause the computing device to:read a dataset that includes a plurality of observation vectors, wherein a unique subset of the plurality of observation vectors is defined for each unique value of a plurality of values of a group variable,
wherein each observation vector of the plurality of observation vectors includes an explanatory variable value of an explanatory variable, a response variable value of a response variable, a group variable value of the group variable, and a factor variable value of a factor variable,
wherein the group variable defines a group associated with a respective unique subset of observation vectors, wherein the factor variable defines a characteristic of each respective group;
(A) train a nonlinear model of a nonlinear model type with each unique subset of the plurality of observation vectors to describe the response variable based on the explanatory variable for each unique value of the plurality of values of the group variable, wherein each trained nonlinear model defines a parameter estimate value for a parameter variable of the nonlinear model type;
(B) present nonlinear model results from the trained nonlinear model defined for each unique value of the plurality of values of the group variable within a first sub-window of a first window of a display;
(C) receive an indicator of a request to perform parameter analysis of the nonlinear model results using the factor variable, wherein the indicator is received in association with the first window of the display when a selector presented in the first window is selected;
(D) train a linear model of the parameter variable of the nonlinear model using the parameter estimate value from each trained nonlinear model with the factor variable value of each unique subset of the plurality of observation vectors as a model effect;
(E) present linear model results from the trained linear model of the parameter variable of the nonlinear model within a second sub-window of the first window of the display; and
(F) present predicted response variable values as a function of the explanatory variable and as a function of the factor variable using the linear model results combined with the nonlinear model results within a third sub-window of the first window of the display, wherein the first window provides an interactive graphical interface to present an effect of the factor variable on the predicted response variable values.

US Pat. No. 10,970,650

AUC-MAXIMIZED HIGH-ACCURACY CLASSIFIER FOR IMBALANCED DATASETS

King Abdulaziz University...

1. An AUC-maximized high-accuracy classification method for imbalanced datasets composed of a majority samples dataset and a minority samples dataset, comprising the steps of:under-sampling of the majority samples dataset to produce k clusters of sub-majority datasets,
combining the minority samples dataset with each of the k clusters of sub-majority datasets to transform an original imbalanced dataset into multiple balanced sub-datasets,
detecting outliers in each balanced sub-dataset,
categorizing the detected outliers into a first category consisting of samples which user input has identified as outliers and a second category of outliers consisting of samples which the user input has not identified as outliers;
removing outliers of the first category from training data and validation data of each balanced sub-dataset,
removing outliers of the second category from only the training data of each balanced sub-dataset,
building an AUC-maximized decision boundary between majority samples and minority samples of each sub-dataset, and
assembling above k MaxAUC classifiers as the final classifier which achieves a higher classification accuracy of the majority samples dataset without changing that of the minority samples dataset.

US Pat. No. 10,970,649

AUTOMATED REINFORCEMENT-LEARNING-BASED APPLICATION MANAGER THAT USES LOCAL AGENTS

VMware, Inc., Palo Alto,...

1. An automated reinforcement-learning-based application manager that manages a computing environment that includes one or more applications, one or more local agents, and one or more of a distributed computing system having multiple computer systems interconnected by one or more networks, a standalone computer system, and a processor-controlled user device, the reinforcement-learning based application manager comprising:one or more processors, one or more memories, and one or more communications subsystems;
a set of actions that can be issued to the computing environment; and
an iterative control process that repeatedly
selects and issues a next action to one or more of the one or more local agents according to a control policy that uses a state vector that represents a current state of the computing system,
receives, from one or more of the one or more local agents in response to execution of the issued next action, one of an observation and state vector, from which the control process generates a current state and a reward, which the control process uses to attempt to learn an optimal or near-optimal control policy.

US Pat. No. 10,970,648

MACHINE LEARNING FOR TIME SERIES USING SEMANTIC AND TIME SERIES DATA

INTERNATIONAL BUSINESS MA...

1. A system, comprising:a memory that stores computer executable components;
a processor that executes computer executable components stored in the memory, wherein the computer executable components comprise:
an analysis component that establishes one or more relationships between one or more elements of semantic data, including one or more time series identifiers, and one or more elements of time series data in a relationship database;
a prediction component that generates one or more advisory outputs, wherein generation of the one or more advisory outputs is performed in response to a trigger event; and
a learning component that determines the one or more relationships in the relationship database, wherein determination of the one or more relationships is based on information indicative of whether the one or more advisory outputs satisfy a defined criterion; and
an adjustment component that adjusts a variable threshold value of the one or more relationships in the relationship database based on the defined criterion, wherein the variable threshold value is a representation of a distance between two or more vectors representing semantic entities or time series data in one or more multidimensional matrices comprising the relationship database, wherein the semantic entities are in a power grid environment,
wherein based on a determination that the one or more advisory outputs satisfy the defined criterion, the system displays a live time series display of temperature in response to a natural language message regarding a defined weather event, and the prediction component outputs time series displays of readings of current.

US Pat. No. 10,970,647

DEEP FEATURE GENERATION FOR CLASSIFICATION

Facebook, Inc., Menlo Pa...

1. A method comprising:by a computing device, obtaining a plurality of sparse features, a plurality of categorical features, and a plurality of numerical features;
by the computing device, generating a plurality of different permutations of features, wherein each permutation of features comprises at least a first feature and a second feature arranged in a specific order that is different from other permutations of features, wherein the first feature is selected from the plurality of sparse features, and wherein the second feature is selected from the plurality of sparse features, the plurality of categorical features, or the plurality of numerical features;
by the computing device, identifying a user account of a social networking system;
by the computing device, generating, for each of the plurality of different permutations of features, one or more deep feature values associated with the user account of the social-networking system, the generating of the one or more deep feature values for the associated permutation of features comprising:
extracting a first feature value associated with the user account using a social graph of the social networking system, wherein the first feature value corresponds to the first feature of the permutation of features;
extracting a second feature value associated with the user account using the extracted first feature value and the social graph of the social networking system, wherein the second feature value corresponds to the second feature of the permutation of features; and
using a combination of the first feature value and the second feature value to generate the one or more deep feature values corresponding to deep features for the permutation of features; and
by the computing device, training a machine-learning model using the one or more deep feature values associated with the user account generated for each of the plurality of different permutations of features, wherein the trained machine-learning model is configured to use deep feature values associated with a second user account and classify whether the second user account is one of a fake user account or a spam user account on the social-networking system based on the deep feature values associated with the second user account.

US Pat. No. 10,970,646

ACTION SUGGESTIONS FOR USER-SELECTED CONTENT

GOOGLE LLC, Mountain Vie...

1. A mobile device comprising:a display device;
at least one processor; and
memory storing instructions that, when executed by the at least one processor, cause the mobile device to:
convert a selection made via the display device into a query,
provide the query to an action suggestion model, the action suggestion model being trained to predict one or more actions given the query, each of the one or more actions being associated with a corresponding mobile application,
receive predicted actions and probability scores that each correspond to a respective one of the predicted actions,
identify metadata about the mobile device, wherein the metadata about the mobile device indicates whether the mobile device is moving,
determine a certain threshold that is dependent on the metadata about the mobile device,
select a subset of the predicted actions based on the subset satisfying the certain threshold, wherein the selected subset includes a lower quantity of mobile applications when the mobile device is moving,
initiate display of the subset of the predicted actions on the display device,
receive a user selection of one of the predicted actions of the subset, and
initiate an intent using the query for the mobile application associated with the user selection.

US Pat. No. 10,970,645

METHOD FOR EXPLAINABLE ACTIVE LEARNING, TO BE USED FOR OBJECT DETECTOR, BY USING BAYESIAN DUAL AUTOENCODER AND ACTIVE LEARNING DEVICE USING THE SAME

Stradvision, Inc., Pohan...

1. A method for explainable active learning, to be used for an object detector, by using a Bayesian dual encoder, comprising steps of:(a) if one or more test images are acquired, an active learning device performing or supporting another device to perform (i) a process of inputting the test images into the object detector, to thereby allow the object detector to detect one or more objects for testing in each of the test images and thus to output bounding boxes for testing corresponding to the objects for testing, (ii) a process of cropping one or more regions, respectively corresponding to the bounding boxes for testing, in each of the test images, to thereby generate first cropped images for testing to n-th cropped images for testing wherein n is an integer equal to or greater than 1, (iii) a process of resizing (1) each of the test images and (2) the first cropped images for testing to the n-th cropped images for testing corresponding to each of the test images into a same size, to thereby generate first resized images for testing to (n+1)-th resized images for testing, and (iv) a process of inputting the first resized images for testing to the (n+1)-th resized images for testing into a data encoder, to thereby allow the data encoder to encode the first resized images for testing to the (n+1)-th resized images for testing and thus to output first data codes for testing to (n+1)-th data codes for testing respectively corresponding to the first resized images for testing to the (n+1)-th resized images for testing;
(b) the active learning device performing or supporting another device to perform (b1) one of (i) (i-1) a process of inputting the test images into the object detector, to thereby allow the object detector to apply m times of Monte Carlo dropouts to each of the test images and thus to generate first feature maps for testing to m-th feature maps for testing corresponding to each of the test images wherein m is an integer equal to or greater than 2, (i-2) a process of applying Bayesian output embedding operation to the first feature maps for testing to the m-th feature maps for testing, to thereby generate activation entropy maps for testing respectively corresponding to the test images, (i-3) a process of cropping one or more regions, corresponding to specific bounding boxes for testing determined as respectively enclosing the objects for testing among the bounding boxes for testing, in each of the activation entropy maps for testing, to thereby generate first cropped activation entropy maps for testing to o-th cropped activation entropy maps for testing wherein o is an integer equal to or less than n and (i-4) a process of resizing (1) each of the activation entropy maps for testing and (2) the first cropped activation entropy maps for testing to the o-th cropped activation entropy maps for testing corresponding to each of the activation entropy maps for testing into a same size, to thereby generate first resized activation entropy maps for testing to (o+1)-th resized activation entropy maps for testing, and (ii) (ii-1) a process of inputting first resized object images for testing to (o+1)-th resized object images for testing, determined as the objects for testing, among the first resized images for testing to the (n+1)-th resized images for testing into the object detector, to thereby allow the object detector to apply m times of the Monte Carlo dropouts to each of the first resized object images for testing to the (o+1)-th resized object images for testing and thus to generate first object feature maps for testing to m-th object feature maps for testing corresponding to each of the first resized object images for testing to the (o+1)-th resized object images for testing and (ii-2) a process of applying the Bayesian output embedding operation to the first feature maps for testing to the m-th feature maps for testing, to thereby generate the first resized activation entropy maps for testing to the (o+1)-th resized activation entropy maps for testing respectively corresponding to the first resized object images for testing to the (o+1)-th resized object images for testing and (b2) a process of inputting the first resized activation entropy maps for testing to the (o+1)-th resized activation entropy maps for testing into a model encoder, to thereby allow the model encoder to encode the first resized activation entropy maps for testing to the (o+1)-th resized activation entropy maps for testing and thus to output first model codes for testing to (o+1)-th model codes for testing corresponding to the first resized activation entropy maps for testing to the (o+1)-th resized activation entropy maps for testing; and
(c) the active learning device performing or supporting another device to perform (i) (i-1) a process of confirming reference data codes corresponding to the number of the resized images equal to or less than a counter threshold by referring to a data codebook, wherein the data codebook is created by referring to previous training images used for pre-training the object detector and wherein the data codebook lists the number of the resized images per data code, (i-2) a process of extracting first specific data codes, respectively corresponding to the reference data codes, from the first data codes for testing to the (n+1)-th data codes for testing, (i-3) a process of selecting first specific test images, corresponding to the first specific data codes, as rare samples to be used for re-training the object detector, and (i-4) a process of updating the data codebook by referring to second specific data codes corresponding to the rare samples, and (ii) (ii-1) a process of confirming reference model codes, corresponding to average entropy values equal to or greater than an entropy threshold, by referring to a model codebook, wherein the model codebook is created by referring to evaluation images before re-training the object detector and wherein the model codebook lists the average entropy values in the resized activation entropy maps per model code of the evaluation images, (ii-2) a process of extracting specific model codes, corresponding to the reference model codes, from the first model codes for testing to the (o+1)-th model codes for testing, and (ii-3) a process of selecting second specific test images, corresponding to the specific model codes, as hard samples to be used for re-training the object detector.

US Pat. No. 10,970,644

ESTIMATING THE CONDITIONAL RESPONSE TIME TO A REQUEST FOR CONFIDENTIAL DATA

Microsoft Technology Lice...

1. A system comprising:a non-transitory machine-readable medium having instructions stored thereon, which, when executed by a processor, cause the system to:
obtain one or more member profiles and corresponding elapsed times indicating, for each of the one or more training member profiles, how long the corresponding member of a social networking service took to respond to a request for confidential data with a submission of confidential data;
extract a first set of one or more features from the one or more training member profiles;
feed the first set of one or more features and corresponding elapsed times into a machine learning algorithm to train a confidential data response time prediction model to output a predicted time to respond to a request for confidential data for a member profile;
obtain a first member profile not included in the one or more training member profiles;
extract a second set of one or more features from the first member profile;
feed the extracted second set of one or more features to the confidential data response time prediction model, outputting the predicted time for a first member corresponding to the first member profile to respond to a request for confidential data;
request first confidential information from the first member;
receive the first confidential information from the first member, the confidential information provided in a first period of time from when it was requested;
compare the first period of time to the predicted time for the first member; and
indicate that the first confidential information is inaccurate based on the difference between the first period of time and the predicted time for the first member.

US Pat. No. 10,970,643

ASSIGNING A FIRE SYSTEM SAFETY SCORE AND PREDICTIVE ANALYSIS VIA DATA MINING

Honeywell International I...

1. A method comprising:identifying one or more pieces of data information from a fire system installed in a particular occupancy region, wherein the one or more pieces of data information include at least one occupation characteristic attribute selected from the group including: a point label, a point type, a detection device obscuration level, a point status history, an occupancy behavior, a software version incompatibility; and at least one functional operation attribute including a fire system trouble status and information related to the fire system's associated failures, selected from: a corrupted configuration database, a hardware error, a secondary power circuit health condition, a condition of a third party system interface to the fire system;
assigning a weighting value to each selected attribute;
calculating a safety score that is based on, at least, the selected at least one occupation characteristic attribute and the at least one functional operation attribute as a function of the selected attributes and the quantity of the weighting value of each attribute;
and
graphically displaying, via a user interface, a visual representation of the safety score including identifying areas of the particular occupancy region impacted by the safety score.

US Pat. No. 10,970,642

DISTINGUISHING HUMAN-DRIVEN DNS QUERIES FROM MACHINE-TO-MACHINE DNS QUERIES

Akamai Technologies, Inc....

1. A non-transitory processor-readable medium having instructions stored thereon, which when executed by one or more processors, cause the one or more processors to implement a method, comprising:receiving a Domain Name System (DNS) query, wherein the DNS query includes a domain name;
applying one or more predetermined rules to the domain name, the outcome of each of the one or more predetermined rules affecting a probability generated by the one or more processors to determine whether to categorize the DNS query as a human-driven DNS query and not a Machine-to-Machine (M2M) DNS query; and,
based on the application of the one or more predetermined rules and the generated probability, categorizing the DNS query as a human-driven DNS query and not a Machine-to-Machine (M2M) DNS query.

US Pat. No. 10,970,641

HEURISTIC CONTEXT PREDICTION ENGINE

State Farm Mutual Automob...

1. A computer-implemented method, executed with processor, comprising:retrieving, by the processor, an un-structured website history transaction data set stored in a first memory;
receiving, by the processor and from a network interface device, an identifier uniquely identifying a user of a plurality of users;
accessing, by the processor, a heuristic algorithm stored in a second memory;
executing the heuristic algorithm, by the processor, and using the un-structured website history transaction data set and the identifier, wherein executing the heuristic algorithm causes the heuristic algorithm to output a correlation score associated with the user;
predicting, by the processor and using the correlation score, a first context indicating a user category;
generating, by the processor and using the first context, a predicted question of the user; and
updating, by the processor and based at least in part on the predicted question, the heuristic algorithm in the second memory using a second context received from the network interface device.

US Pat. No. 10,970,640

DETERMINING A RISK SCORE USING A PREDICTIVE MODEL AND MEDICAL MODEL DATA

International Business Ma...

1. A computer-implemented method for text analysis of medical study data to extract predictive data, comprising:performing natural language processing on a document in a collection of documents to determine whether the document contains medical model data;
in response to determining that the document contains medical model data, generating annotations for content relating to the medical model data in the document using section and modeling rule dictionaries;
for a first section of the document, identifying one or more first medical modeling rules in the first section using the annotations, wherein a medical modeling rule of the first medical modeling rules is generated using a predictor factor annotation value, a range feature, and a measurement unit;
generating a first medical model based on the annotations for the identified one or more first medical modeling rules and a certainty threshold;
in response to the certainty threshold meeting a user setting, adding the first medical model to a predictive model;
for a second section of the document, identifying one or more second medical modeling rules in the second section using the annotations;
updating the predictive model based on the identified one or more second medical modeling rules and the certainty threshold; and
using the predictive model against unstructured or structured content to generate a predictive score.

US Pat. No. 10,970,639

COGNITIVE ROBOTICS ANALYZER

Accenture Global Solution...

1. A computer-implemented method comprising:receiving first action data that identifies one or more first actions and second action data that identifies one or more second actions;
receiving first process data that identifies a first process performed by the one or more first actions and second process data that identifies a second process performed by the one or more second actions;
receiving process similarity data that indicates that the first process is similar to the second process;
training, using the first action data, the second action data, additional action data, the first process data, the second process data, additional process data, the process similarity data, and additional process similarity data, the predictive model that is configured to classify one or more actions as performing a process that is similar or not similar to the first process and the second process;
receiving data that identifies a group of actions that perform the first process and that include an unnecessary action for the group of actions of the process;
applying, to the group of actions, a predictive model that is configured to classify one or more actions as performing the process that is similar or not similar to the first process;
based on applying the predictive model to the one or more of the actions, classifying the process performed by the group of actions as similar to the first process;
identifying, from among the group of actions, a particular action;
applying the predictive model to the group of actions with the particular action removed;
based on applying the predictive model to the group of actions with the particular action removed, classifying the process performed by the group of actions with the particular action removed as similar to the first process; and
based on classifying the process performed by the group of actions as similar to the first process and based on classifying the process performed by the group of actions with the particular action removed as similar to the first process, determining that the particular action is the unnecessary action for the group of actions; and
based on determining that the particular action is the unnecessary action for the group of actions, providing guidance to a user that performed the unnecessary action.

US Pat. No. 10,970,638

PERSONALLY IDENTIFIABLE INFORMATION IDENTIFICATION

AMERICAN EXPRESS TRAVEL R...

1. A method, comprising:identifying, by a processor, personally identifiable information (PII) within a data model based on processing rules, to create identified PII, wherein the data model comprises entity information about an entity;
comparing, by the processor, the identified PII with established PII in a data bank;
validating, by the processor, the identified PII in response to the identified PII matching the established PII, to create validated PII;
marking, by the processor, the validated PII with a PII marker in response to the validating the identified PII;
detecting, by the processor, that the identified PII is different than the established PII to create differing PII;
adding, by the processor, the differing PII to the data bank; and
associating, by the processor, the differing PII with the established PII.

US Pat. No. 10,970,637

ANALYTICS GATHERING OF DISCOVERED AND RESEARCHED INSIGHTS

INTERNATIONAL BUSINESS MA...

1. A computer-implemented method for optimizing research of an abstracted issue with a plurality of analytics engines, the method comprising:receiving, via an analytics engine controller, a problem report comprising one or more symptoms of a problem in a computing system;
forwarding the problem report via the analytics engine controller to a research optimization engine;
abstracting, via the research optimization engine, one or more issues associated with the problem based on the one or more symptoms of the problem;
obtaining, via the research optimization engine, anomaly research data about an anomaly for one or more of diagnosing the problem and fixing the problem, wherein the anomaly research data is based on the one or more abstracted issues;
associating, via the research optimization engine, the one or more abstracted issues with one or more portions of the anomaly research data; and
assigning, via the research optimization engine, the one or more abstracted issues and the one or more associated portions of the anomaly research data to at least one of the plurality of analytics engines.

US Pat. No. 10,970,636

ADAPTIVE SYSTEM FOR PREDICTING INTEGRATION FAILURES IN LARGE SYSTEMS

Intuit Inc., Mountain Vi...

1. A computer-implemented method for predicting an integration failure, the method comprising:monitoring a plurality of integration builds between source components of a software application under development, wherein each integration build of the plurality of integration builds includes at least a first one of the source components and a second one of the source components, and wherein the monitoring includes determining whether each respective integration build of the monitored plurality of integration builds was successful;
performing a comparison of the monitored plurality of integration builds with prior integration builds of the software application, wherein performing the comparison comprises:
identifying an integration pattern between the first one of the source components and the second one of the source components based on the monitored plurality of integration builds;
identifying a prior successful integration pattern;
determining a minimum distance measure between the integration pattern and the prior successful integration pattern; and
determining the prior successful integration pattern is not identical to the integration pattern based on the minimum distance measure;
predicting, based on the comparison, a probability of successfully building the software application from the first one of the source components and second one of the source components;
determining, based on the comparison, one or more actions to be taken with respect to a given source component of the source components to increase the probability of successfully building the software application from the source components;
determining, based on the comparison, a particular developer to perform the one or more actions by determining that the particular developer worked on the given source component; and
notifying the particular developer that the one or more actions should be performed.

US Pat. No. 10,970,635

DATA PROCESSING FOR MAKING PREDICTIVE DETERMINATIONS

1. A system comprising:a memory comprising computer-executable instructions; and
a processor configured to access the memory to execute the computer-executable instructions to collectively at least:
monitor a data stream comprising a plurality of messages to detect a conclusion event associated with a particular message of the plurality of messages, the particular message being a standard message type for carrying information relating to the conclusion event;
determine, based on the conclusion event, that a dependent user has been released from a facility and that care related to presence of the dependent user at the facility has concluded;
access, based on the particular message, a release summary associated with a record of the dependent user, at least a portion of the release summary comprising unstructured data prepared by an authorized user who responded to conditions of the dependent user while the dependent user was present at the facility;
identify, based on the release summary, one or more related reports generated in connection with the care related to the presence of the dependent user at the facility or other care unrelated to the presence of the dependent user at the facility, the one or more related reports comprising other unstructured data;
evaluate at least one of the unstructured data of the release summary or the other unstructured data of the one or more related reports using a natural-language processing model to generate a list of potential follow-up tasks; and
generate a follow-up plan for execution of the list of potential follow-up tasks, the follow-up plan comprising:
a ranking of the list of potential follow-up tasks according to prioritization of medical necessity; and
a contact plan for contacting the dependent user to schedule one or more follow-up appointments corresponding to at least one potential follow-up task of the list of potential follow-up tasks.

US Pat. No. 10,970,634

METHODS AND SYSTEMS FOR CAPTURING ANALYTIC MODEL AUTHORING KNOWLEDGE

General Electric Company,...

1. A computer system implementing a development environment for generating predictive models using model parameters provided by a predictive model authoring tool, the computer system comprising:the predictive model authoring tool configured to:
develop a predictive analytic via a development interface of the predictive model authoring tool based on one or more user modeling task inputs provided to interface controls of the predictive model authoring tool;
determine a modeling context for the predictive analytic based on modeling operations that are used to develop the predictive analytic;
provide the modeling context, the one or more user modeling task inputs, and one or more model parameters of the predictive analytic to a model development context analyzer which determines a related element of an ontology related to the predictive analytic;
query a database using the at least one element of the ontology, wherein the querying returns derived modeling knowledge including one or more recommended model parameters to be added to a machine learning model of the predictive analytic based on at least one of the provided modeling context, the one or more user modeling task inputs, or the one or more model parameters; and
providing the one or more recommended model parameters of the machine learning model of the predictive analytic to the predictive model authoring tool.

US Pat. No. 10,970,633

METHOD FOR OPTIMIZING ON-DEVICE NEURAL NETWORK MODEL BY USING SUB-KERNEL SEARCHING MODULE AND DEVICE USING THE SAME

STRADVISION, INC., Gyeon...

1. A method for optimizing an on-device neural network model by using a Sub-kernel Searching Module (SSM), comprising steps of:(a) a learning device, on condition that a Big Neural Network Model (B-NN Model) having a maximal capacity capable of performing a targeted task by using a maximal computing power allowed by an edge device on which a neural network model is to be mounted has been trained to generate a first inference result on an input data fed thereinto, if a training data is acquired, performing or supporting another device to perform a process of inputting the training data into a Sub-kernel Searching Module (SSM), to thereby allow the Sub-kernel Searching Module to (i) identify at least one constraint corresponding to the training data wherein the constraint is imposed on at least part of a computational load, parameters, a latency, an accuracy, and flops, (ii) generate a state vector corresponding to the training data wherein the state vector includes information on at least one of a degree of complexity of the training data and a degree of sharpness of the training data, and (iii) generate architecture information on a specific sub-kernel having a minimal kernel size and a minimal number of channels suitable for performing the targeted task on the training data having the state vector under the constraint, wherein the specific sub-kernel is a subset of a super kernel corresponding to the maximal capacity of the Big Neural Network Model and is comprised of a kernel size equal to or less than that of the super kernel and the number of the channels in the kernel equal to or less than that in the super kernel;
(b) the learning device performing or supporting another device to perform (i) a process of optimizing the Big Neural Network Model by referring to the architecture information on the specific sub-kernel, to thereby generate a specific Small Neural Network Model (S-NN Model) having the architecture information on the specific sub-kernel, from the Big Neural Network Model and (ii) a process of allowing the specific Small Neural Network Model to generate a second inference result on the training data; and
(c) the learning device performing or supporting another device to perform a process of calculating one or more first losses by referring to the first inference result and the second inference result and a process of training the Sub-kernel Searching Module by using the first losses.

US Pat. No. 10,970,632

GENERATING A SCORE FOR A RUNBOOK OR A SCRIPT

HCL TECHNOLOGIES LTD, Ut...

1. A system for generating a score for one of a runbook or a script, the system comprising:a memory;
a processor coupled to the memory, wherein the processor is configured to execute programmed instructions stored in the memory to:
receive ticket data, wherein the ticket data comprises a ticket, and a ticket description;
identify a set of policies associated with a platform, wherein the set of policies is identified based on an analysis of historical data, the ticket and the ticket description;
determine a set of functions based on an analysis of the set of policies, wherein the set of policies are analysed using a neural network technique and an Inverse Reinforcement Learning technique;
recommend one of a runbook or a script from a runbook recommendation engine based on an analysis of the ticket and the ticket description, wherein the ticket and the ticket description are analysed using the set of functions;
identify a set of hyperparameter values associated with the platform, wherein the set of hyperparameter values is further used to implement the runbook recommendation engine over a new platform, and wherein a transfer learning technique is used for the implementation of the runbook recommendation engine;
execute the runbook or the script to resolve the ticket;
record one of a script success or a script failure based on the execution of the runbook or the script, wherein the script success is recorded when the runbook or the script resolves the ticket, and wherein the failure script is recorded when the runbook or the script fails to resolve the ticket; and
generate a score associated with one of the runbook or the script, wherein the score is generated based on one of the script success or the script failure.

US Pat. No. 10,970,631

METHOD AND APPARATUS FOR MACHINE LEARNING

AUTOCRYPT CO., LTD., Seo...

1. A method of machine learning for a convolutional neural network (CNN), the method comprising:receiving input target data;
determining whether to initiate incremental learning on the basis of a difference between a statistical characteristic of the target data with respect to the CNN and a statistical characteristic of previously used training data with respect to the CNN:
determining a set of kernels with a high degree of mutual similarity in each of convolution layers included in the CNN when the incremental learning is determined to be initiated;
updating each kernel included in the set of kernels by calculating a random matrix with each of the kernels included in the set of kernels with the high degree of mutual similarity; and
updating a weight between nodes to which kernels included in the set of kernels with a high degree of mutual similarity are applied,
wherein the determining of the set of kernels with the high degree of mutual similarity comprises:
constructing concatenated kernel matrices from kernels used in a convolution operation of the node;
measuring similarity between the concatenated kernel matrices based on an absolute value of a difference between the concatenated kernel matrices; and
determining a pair of the kernels with the smallest absolute value of the difference between the concatenated kernel matrices, wherein the pair of the kernels belongs to the set of the kernels,
wherein determining of whether to initiate the incremental learning comprises when the target data is a data array consisting of a plurality of pieces of data, it is determined that the incremental learning is initiated when the number of pieces of data, which are included in the data array and satisfy a condition in which a maximum value among output values of the CNN is less than the predetermined reference value, is greater than or equal to a reference number,
when it is determined that the incremental learning is initiated determining a set of weight vectors with a high degree of mutual similarity in each fully connected layer included in the CNN; and
updating a weight between nodes to which weight vectors included in the set of weight vectors with a high degree of mutual similarity;
wherein the determining of the set of weight vectors with a high degree of mutual similarity comprises measuring at least one pair of the weight vectors with a high degree of mutual similarity by measuring a distance or similarity between weight vectors.

US Pat. No. 10,970,630

NEUROMORPHIC COMPUTING ARCHITECTURE WITH DYNAMICALLY ACCESSIBLE CONTEXTS

Lewis Rhodes Labs, Inc., ...

1. A neuromorphic computing system comprising:a plurality of neuron circuits;
a plurality of synapse connections, the synapse connections connecting each of the neuron circuits to at least one other neuron circuit in the neuron circuits; and
a resource allocation system that allocates resources of the neuromorphic computing system to execute a neural network algorithm, the resource allocation system configured to perform acts comprising:
identifying a plurality of subgraphs of a neural algorithm graph, the neural algorithm graph representative of operations of the neural network algorithm; and
assigning execution of the subgraphs to a first group of neuron circuits in the neuron circuits, wherein the first group of neuron circuits executes the subgraphs serially, wherein execution of each of the subgraphs results in generation of output data, the first group of neuron circuits configured to provide the output data for each of the subgraphs to a second group of neuron circuits in the neuron circuits, wherein the neuron circuits are configured such that the second group of neuron circuits receives the output data for each of the subgraphs at a same time, the second group of neuron circuits configured to perform one or more processing operations based upon the output data.

US Pat. No. 10,970,629

ENCODINGS FOR REVERSIBLE SPARSE DIMENSIONALITY REDUCTION

Amazon Technologies, Inc....

1. A system comprising:computer-readable memory storing executable instructions; and
one or more hardware-based processors programmed by the executable instructions to at least:
obtain training data comprising a training data input vector and a corresponding reference data output vector, wherein the training data input vector comprises a first quantity of values, and wherein the reference data output vector comprises a second quantity of values;
generate an encoded training data input vector using the training data input vector and a plurality of mapping functions of a probabilistic data structure, wherein the encoded training data input vector comprises a representation of the training data input vector having a third quantity of values that is less than the first quantity of values;
generate an encoded reference data output vector using the corresponding reference data output vector and a plurality of mapping functions of the probabilistic data structure, wherein the encoded reference data output vector comprises a representation of the corresponding reference data output vector having a fourth quantity of values that is less than the second quantity of values;
obtain an artificial neural network configured to receive, as input, an input vector comprising the third quantity of values and generate, as output, an output vector comprising the fourth quantity of values;
generate an encoded training data output vector using the artificial neural network and the encoded training data input vector;
compute a difference between the encoded training data output vector and the encoded reference data output vector;
compute a gradient based on the difference between the encoded training data output vector and the encoded reference data output vector;
update a parameter of the artificial neural network using the gradient;
obtain an input vector;
generate an encoded input vector using the input vector, wherein the encoded input vector comprises a compressed representation of the input vector having fewer values than the input vector;
generate an encoded output vector using the artificial neural network and the encoded input vector, wherein the encoded output vector comprises compressed representation of an output vector having fewer values than the encoded output vector;
generate a first value of a value space using a product of a subset of values of the encoded output vector; and
compute a probability of purchase for an item using the first value.

US Pat. No. 10,970,628

TRAINING NEURAL NETWORKS REPRESENTED AS COMPUTATIONAL GRAPHS

Google LLC, Mountain Vie...

1. A method for training a neural network represented as a computational graph, wherein the computational graph comprises a plurality of nodes, a plurality of connector directed edges, and a plurality of parameter directed edges, wherein each node represents a respective operation performed by the neural network as part of determining a neural network output from a neural network input, wherein each connector directed edge connects a respective first node to a respective second node that represents an operation that receives, as input, an output of an operation represented by the respective first node, and wherein each parameter directed edge connects into a respective node and represents a flow of one or more parameters of the neural network as input to the operation represented by the respective node, the method comprising:obtaining data representing the computational graph;
augmenting the computational graph to generate a training computational graph for training the neural network using a machine learning training algorithm that includes computing a gradient of an objective function with respect to each of the parameters of the neural network, comprising:
identifying one or more control flow nodes in the computational graph, wherein each control flow node represents a control flow operation that causes operations represented by one or more other nodes in the computational graph to be conditionally or iteratively performed during execution of the computational graph, wherein the one or more control flow nodes comprise (i) an enter control flow node representing an enter operation and (ii) a switch control flow node representing a switch operation,
inserting a plurality of gradient nodes and training edges into the computational graph to generate a backward path through which gradient data flows, the backward path representing operations for computing the gradients of the objective function with respect to the parameters of the neural network, wherein each gradient node represents a gradient function that computes a gradient of the objective function with respect to parameters flowing along a respective parameter directed edge in the computational graph,
inserting, for each identified control flow node, a corresponding backward path control flow node along the backward path through the computational graph, the corresponding backward path control flow node representing an operation for adjusting how the gradient data flows through the gradient nodes along the backward path to account for a respective operation represented by the identified control flow node, wherein a corresponding backward path control flow node for the enter control flow node is an exit control flow node representing an exit operation, wherein a corresponding backward path control flow node for the switch control flow node is a merge backward control flow node representing a merge operation along the backward path; and
training the neural network using the machine learning training algorithm by executing the training computational graph.

US Pat. No. 10,970,627

TIME BORROWING BETWEEN LAYERS OF A THREE DIMENSIONAL CHIP STACK

Xcelsis Corporation, San...

1. A three-dimensional (3D) circuit comprising:a first integrated circuit (IC) die comprising a first circuit;
a second IC die comprising a second circuit, the second IC die vertically stacked with the first IC die; and
at least one latch to transfer signals between the first circuit on the first IC die and the second circuit on the second IC die through a direct-bonded connection between the first and second IC dies that crosses a bonding layer that bonds the first and second IC dies, the latch operating in an open first mode for allowing a signal to pass from the first circuit to the second circuit and a closed second mode for maintaining the signal passed through during the prior open first mode.

US Pat. No. 10,970,626

MULTI-MEMRISTIVE SYNAPSE WITH CLOCK-ARBITRATED WEIGHT UPDATE

International Business Ma...

1. A computer-implemented method for providing a multi-memristive synaptic element for a cognitive computing system, said multi-memristive synaptic element comprising an array of memristive devices, said method comprising:arbitrating, by a programmed hardware processor, a synaptic weight allocation, a related synaptic weight being represented by a synaptic weight variable of said multi-memristive synaptic element,
updating, using said programmed hardware processor, said synaptic weight variable by a delta amount, and
assigning, by the programmed hardware processor, said memristive devices to elements of a clock-like ordered circular list for selecting a particular memristor of said memristive devices requiring to be updated by a deterministic, periodic global clock that points to a different memristor at every clock tick,
running, by said programmed hardware processor, a reset clock, operating independently of said periodic global clock, for triggering a reset operation to be performed on a memristive device;
such that said multi-memristive synaptic element has a larger dynamic range and a more linear conductance response than a single memristor synaptic element; and
wherein a number of said ordered list of elements in said ordered circular list and an increment rate for said ordered circular list relate to each other in such a way that during consecutive iterations of selections at least one sub-element is different.

US Pat. No. 10,970,625

DEVICE WITH MULTIPLE RESISTANCE SWITCHES WITH DIFFERENT SWITCHING CHARACTERISTICS

Hewlett Packard Enterpris...

1. A device comprising:a crossbar array comprising a plurality of memristor-based spatio-temporal (MST) synapses, wherein a respective MST synapse provides spatial and temporal weighting functions, and wherein the MST synapse comprises:
a first resistance switch comprising a first memristor connected in parallel with a resistor; and
a second resistance switch comprising a second memristor and connected in series with the first resistance switch;
wherein the first and second resistance switches have different switching voltages to switch between on and off states.

US Pat. No. 10,970,624

PRE-PROGRAMMED RESISTIVE CROSS-POINT ARRAY FOR NEURAL NETWORK

Samsung Electronics Co., ...

1. A semiconductor device comprising:a base insulating layer;
a first set of conductive contacts for a first resistive device, the first set of conductive contacts extending through the base insulating layer;
a second set of conductive contacts for a second resistive device, the second set of conductive contacts extending through the base insulating layer;
a semiconducting layer deposited on a top surface of the base insulating layer, a first surface of the semiconducting layer in contact with the first set and the second set of conductive contacts;
a top insulating layer deposited on a second surface and opposite sidewalls of the semiconducting layer, the top insulating layer comprising at least two portions that expose the second surface of the semiconducting layer, a first portion and a second portion, wherein a length of the first portion along the exposed second surface of the semiconducting layer is different from a length of the second portion, wherein the first portion is associated with the first resistive device and the second portion is associated with the second resistive device, and wherein a bottom surface of the top insulating layer is formed directly on the top surface of the base insulating layer; and
a shunting layer deposited directly on the second surface of the semiconducting layer, the shunting layer comprising a first portion having a first area and a second portion having a second area different than the first area, the first portion isolated from the second portion by the top insulating layer, the shunting layer being deposited inside a removed portion of the top insulating layer without penetration above a top surface of the top insulating layer, a thickness of the shunting layer being equal to a thickness of the top insulating layer, and the shunting layer configured to reduce a resistance between a first contact and a second contact of the first set of conductive contacts across the semiconducting layer.

US Pat. No. 10,970,623

SYSTEM AND METHOD FOR TRAINING ARTIFICIAL INTELLIGENCE SYSTEMS USING A SIMA BASED PROCESSOR

Alphaics Corporation, Re...

1. A system comprising:a first memory module configured to store a plurality of reinforcement learning instructions that implement one or more reinforcement learning operations;
an instruction fetch and decode unit operably coupled to the first memory module, wherein the instruction fetch and decode unit is configured to decode each of the plurality of reinforcement learning instructions to generate a plurality of threads; and
a reinforcement learning processor operably coupled to the instruction fetch and decode unit, wherein the reinforcement learning processor is configured to apply each of the plurality of threads in parallel to at least one of a reinforcement learning agent and a reinforcement learning environment to generate an output corresponding to each of the plurality of threads, and wherein the reinforcement learning processor is further configured to train the reinforcement learning agent by optimizing a neural network based upon the outputs.

US Pat. No. 10,970,622

DYNAMIC GATING USING NEUROMORPHIC HARDWARE

INTERNATIONAL BUSINESS MA...

1. A system comprising:a neurosynaptic core, the neurosynaptic core comprising a plurality of neurons and axons, wherein
the neurosynaptic core comprises a programmable gate operative to receive a control signal and selectively output a first output signal based on the control signal, the control signal comprising open and close control spikes, the programmable gate configured to count the open and close spikes and to output the first output signal only when a difference between the count of open and close control spikes exceeds a predetermined threshold;
the neurosynaptic core is operative to receive a first input signal, and wherein the first output signal comprises the first input signal,
the first input signal comprises a plurality of spikes, and
the first input signal is rate coded.

US Pat. No. 10,970,621

METHODS TO PREDICT FOOD COLOR AND RECOMMEND CHANGES TO ACHIEVE A TARGET FOOD COLOR

NOTCO DELEWARE, LLC, San...

1. A computer-implemented method to determine color of a food item, the method comprising:generating, by a first model, multi-dimensional embedding vectors from a plurality of ingredients;
receiving an initial formula of an item, the initial formula comprising a first set of ingredients and its respective quantities;
representing the initial formula as a sparse proportional vector comprising a respective dimension for each of the first set of ingredients;
generating an ingredient embedding vector from the sparse proportional vector by using the multi-dimensional embedding vectors;
extracting a color from an image of the item;
encoding the color for representing in a color space;
configuring a second model given the ingredient embedding vector and the color, wherein the configuring includes implementing corrections to the second model for the color represented in the color space;
predicting a particular color of a given formula based on the corrections and the multi-dimensional embedding vectors;
applying a third model to the particular color of the given formula to determine a recommended formula, the recommended formula including changes to the given formula to achieve a target color, wherein the third model implements at least one loss function to determine the changes to the given formula, wherein the at least one loss function considers a plurality of factors including:
an attribute distance between the given formula and the target color,
an ingredient distance between the given formula and recommended formula,
an ingredient likelihood of the recommended formula to be a realistic formula, and
an ingredient sparsity of ingredients in the recommended formula.

US Pat. No. 10,970,620

OPTIMIZING PRODUCTION USING DESIGN OF EXPERIMENT AND RESERVOIR MODELING

Halliburton Energy Servic...

1. A method comprising:forming a data analytics model of a hydrocarbon reservoir, wherein the data analytics model produces a significant input parameter;
building a historical deep recursive neural network (RNN) model based on time-series production data from the hydrocarbon reservoir;
storing the historical deep RNN model on a data storage device;
designing an experiment to produce a significant experimental input parameter;
inputting an output from the historical deep RNN into the experiment;
running the experiment to produce the significant experimental input parameter;
comparing the significant input parameter to the significant experimental input parameter stored in the data analytics model to produce a difference between the significant input para meter and the significant experimental input parameter; and
adjusting the significant experimental input parameter of the data analytics model to reduce the difference.

US Pat. No. 10,970,619

METHOD AND SYSTEM FOR HIERARCHICAL WEIGHT-SPARSE CONVOLUTION PROCESSING

MOFFETT TECHNOLOGIES CO.,...

1. A computer-implemented method, comprising:obtaining an input tensor and a plurality of filters at a convolution layer of a neural network;
segmenting the input tensor into a plurality of sub-tensors and assigning the plurality of sub-tensors to a plurality of processors;
generating, for each of the plurality of filters, a hierarchical bit representation of a plurality of non-zero weights in the filter, wherein
the hierarchical bit representation comprises a first layer and a second layer,
the first layer comprises a plurality of bits corresponding to a plurality of sub-filters in the filter and indicating whether the corresponding sub-filter has at least one non-zero weight, and
the second layer comprises a plurality of key-value pairs corresponding to the plurality of non-zero weights in the filter, each key-value pair comprising one of the plurality of non-zero weights as value and a channel identifier of the one of the plurality of non-zero weights as key;
identifying, by each of the plurality of processors based on the hierarchical bit representation, one or more of the plurality of non-zero weights and corresponding input values from the assigned sub-tensor;
performing multiply-and-accumulate (MAC) operations on the one or more identified non-zero weights and the corresponding input values to obtain a partial sum; and
accumulating the partial sum obtained by each of the plurality of processors to construct an output tensor of the convolution layer of the neural network.

US Pat. No. 10,970,618

LOGICAL ENTANGLEMENT DEVICE FOR GOVERNING AI-HUMAN INTERACTION

1. A logic entanglement device, comprising:at least one processor;
at least one memory;
at least one biometric sensor coupled to at least one biological organism;
at least one diagnostic sensor coupled to at least one machine; and
a transmitter coupled to the at least one biometric sensor and the at least one diagnostic sensor,
wherein the logic entanglement device receives a plurality of functionality assessments comprising biological organism functionality assessments generated via the at least one biometric sensor and a machine functionality assessments generated via the at least one diagnostic sensor and transmits an output signal based on the content of the functionality assessments.

US Pat. No. 10,970,617

DEEP CONVOLUTIONAL NEURAL NETWORK ACCELERATION AND COMPRESSION METHOD BASED ON PARAMETER QUANTIFICATION

INSTITUTE OF AUTOMATION C...

1. An acceleration and compression method for a deep convolutional neural network based on quantization of a parameter, comprising:quantizing the parameter of the deep convolutional neural network to obtain a plurality of subcode books and respective corresponding index values of the plurality of subcode books;
acquiring an output feature map of the deep convolutional neural network according to the plurality of subcode books and respective corresponding index values of the plurality of subcode books;
the deep convolutional neural network comprises a plurality of convolutional layers, the quantizing a parameter of the deep convolutional neural network to obtain a plurality of subcode books and respective corresponding index values of the plurality of books comprises:repeatedly executing the following processes until the plurality of convolutional layers have all been traversed:dividing a parameter of the convolutional layers into M groups, each group comprising a plurality of first parameter sub-vectors, wherein, M is a positive integer;
performing a K-means clustering for the plurality of first parameter sub-vectors of each group to obtain a corresponding first subcode book of each group, the first subcode book comprising a plurality of first code words;
calculating Euclidean distances between the plurality of first parameter sub-vectors and the plurality of first code words, respectively;
acquiring a first parameter sub-vector and a first code word with a minimum of the Euclidean distances, and recording an index value of the first code word with the minimum of Euclidean distances;
the deep convolutional neural network comprises a plurality of fully connected layers, the quantizing a parameter of the deep convolutional neural network to obtain a plurality of subcode books and respective corresponding index values of the plurality of subcode books comprises:
repeatedly executing the following processes until the plurality of fully connected layers have all been traversed:
dividing a parameter of the fully connected layers into M groups, each group comprising a plurality of second parameter sub-vectors;
performing a K-means clustering for the plurality of second parameter sub-vectors respectively to obtain a corresponding second subcode book of each of the groups, the second subcode book comprising a plurality of second code words;
calculating Euclidean distances between the plurality of second parameter sub-vectors and the plurality of second code words, respectively;
acquiring a second parameter sub-vector and a second code word with a minimum of the Euclidean distances, and recording an index value of the second code word with the minimum of Euclidean distances;
the deep convolutional neural network comprises the plurality of convolutional layers, the acquiring an output feature map of the deep convolutional neural network according to the plurality of subcode books and respective corresponding index values of the plurality of subcode books comprises:
repeatedly executing the following processes until the plurality of convolutional layers have all been traversed:
inputting a first feature map;
dividing the first feature map into M groups, each of the groups comprising a plurality of first feature sub-vectors;
performing inner product operations between the plurality of first feature sub-vectors and the plurality of first code words in a corresponding first subcode book of each group respectively to obtain a plurality of first operation results, and storing the plurality of the first operation results in a first lookup table;
summing the plurality of first operation results to obtain an output first feature map.

US Pat. No. 10,970,615

SYSTEMS AND METHODS FOR GENERATING SECURE TAGS

Laava ID Pty Ltd, Sydney...

1. An authentication server for interacting with secure tags, comprising:at least one processor; and
at least one non-transitory memory containing instructions that, when executed by the at least one processor, cause the authentication server to perform operations comprising:
receiving a tag identification request including a tag image;
identifying a secure tag in the received tag image using a stored hash of an image of a secure tag generated using a stylesheet;
generating tag data using the received tag image and decoding rules for decoding tags generated using the stylesheet, wherein the decoding rules include:
first decoding rules enabling decoding of at least one of:
a first portion of the secure tag defined by the stylesheet, or
a first subset of tag features defined by the stylesheet;
second decoding rules enabling decoding of at least one of:
a second portion of the secure tag, or
a second subset of the tag features; and
wherein generating the tag data using the received tag image and the decoding rules comprises:
generating first tag data using the first decoding rules; and
generating second tag data using the first tag data and the second decoding rules; and
providing the tag data in response to the tag identification request.

US Pat. No. 10,970,614

SINGLE-USE PRESSURE TRANSDUCER DISPOSABLE INTERFACE

Rosemount Inc., Shakopee...

1. A single-use adapter for coupling a single-use container to a reusable sensor transducer, the single-use adapter comprising:an attachment region;
a deflectable diaphragm sealingly coupled to the attachment region and configured to contact a media sample;
a radio-frequency identification (RFID) tag within the single-use adapter and configured to store and transmit data; and
wherein the RFID tag comprises an open antenna circuit, with exposed contacts, configured to be closed upon physically coupling the adapter to the sensor transducer, wherein, upon closing, the antenna circuit, the antenna circuit is configured to receive radio frequency energy from an RFID reader.

US Pat. No. 10,970,613

SYSTEMS AND METHODS FOR PROVIDING TAGS ADAPTED TO BE INCORPORATED WITH OR IN ITEMS

Sensormatic Electronics, ...

20. A system, comprising:at least one device that:
dynamically determines, during a fabrication of an item, a length of each metal thread to be incorporated into an item to optimize tag performance in view of dielectric and tuning properties of the item;
creates, during the fabrication of the item, at least one metal thread having the length;
sews, during the fabrication of the item, the at least one metal thread into the item being produced to form at least one antenna for a first tag; and
attaches at least a communications enabled device after sewing the at least one metal thread to the item so as to form an electrical coupling or connection between the communications enabled device and the at least one antenna.

US Pat. No. 10,970,612

INTERACTIVE CORE FOR ELECTRONIC CARDS

FITEQ, INC., Lakeland, F...

1. An electronic core for an electronic card, comprising:a rear adhesive layer configured to form a rear top planar surface and a rear bottom planar surface;
a front adhesive layer configured to form a front top planar surface and a front bottom planar surface; and
an interactive core configured within the rear top planar surface and the front bottom planar surface, wherein said interactive core is comprised of:
a stiffening spacer configured with an interior opening surrounded by an outer edge border located between the rear top planar surface and the front bottom planar surface;
a printed circuit board (“PCB”) configured within the interior opening;
an initialization antenna which is configured for activating and personalizing the electronic core and is not configured for use with payment transactions; and
a thermosetting polymeric material configured within the interior opening between the rear top planar surface and the front bottom planar surface;
wherein the electronic core is configured so that it has an outer surface shape defined by the rear bottom planar surface, the front top planar surface and the outer edge border.

US Pat. No. 10,970,611

DOUBLE-SIDED BUTTON WITH A TOUGH FUNCTION

1. A push-button comprising:a first switch contact and a second switch contact which are arranged spaced apart and parallel to each other,
wherein the first switch contact is designed for approaching the second switch contact until the first switch contact touches the latter and for establishing an electrical line connection to the second switch contact when force is applied,
wherein:
the push-button comprises a third switch contact which is arranged spaced apart from the second switch contact and in parallel on a side opposite to the first switch contact,
wherein the third switch contact is designed for approaching the second switch contact until the third switch contact touches the latter and for establishing an electrical line connection to the second switch contact when force is applied,
and the push-button comprises an evaluation unit which is designed for additionally operating the first and/or the third switch contact(s) as a capacitive or inductive touch sensor.

US Pat. No. 10,970,610

RFIC MODULE AND RFID TAG

MURATA MANUFACTURING CO.,...

1. An RFIC module comprising:a substrate;
an RFIC mounted on or in the substrate;
RFIC side first and second terminal electrodes each connected to the RFIC;
antenna side first and second terminal electrodes that are formed on or in the substrate and that are each directly connected or capacitively coupled to an antenna; and
an impedance matching circuit formed on or in the substrate and connected to the RFIC side first and second terminal electrodes and the antenna side first and second terminal electrodes,
wherein the impedance matching circuit includes a first inductor, a second inductor, a third inductor, a fourth inductor, and a fifth inductor,
wherein the first inductor is connected between the antenna side first terminal electrode and the RFIC side first terminal electrode,
wherein the second inductor is connected between the antenna side second terminal electrode and the RFIC side second terminal electrode,
wherein the third inductor has a first end connected to the antenna side first terminal electrode,
wherein the fourth inductor has a first end connected to the antenna side second terminal electrode,
wherein the fifth inductor is connected between a second end of the third inductor and a second end of the fourth inductor,
wherein the first, second, third and fourth inductors each comprise a conductor pattern of a coil shape wound along a surface of the substrate, and the fifth inductor comprises a conductor pattern of an unwound shape,
wherein the first and third inductors are respectively formed in different layers of the substrate and are positioned to have coil openings that overlap each other,
wherein the second and fourth inductors are respectively formed in different layers of the substrate and are positioned to have coil openings that overlap each other, and
wherein the second inductor and the fourth inductor, and the first inductor and the third inductor are arranged in such a positional relationship as to interpose a mounting position of the RFIC along the surface of the substrate.

US Pat. No. 10,970,609

TAG MANAGEMENT DEVICE, TAG MANAGEMENT METHOD, AND PROGRAM

NEC CORPORATION, Tokyo (...

1. A tag management device comprising:an antenna configured to output a readout signal that instructs an old RFID tag and a new RFID tag to emit signals, and receive signals emitted in response to the readout signal from the old RFID tag and the new RFID tag;
a signal detecting unit configured to detect signals which are emitted from the old RFID tag and the new RFID tag;
a comparison unit configured to compare, as first comparison process, IDs included in the signals emitted from the old RFID tag and the new RFID tag;
a search managing unit configured to instruct, as a search mode process, the antenna to continue outputting the readout signal when the IDs included in signals emitted from the old RFID tag and the new RFID tag do not coincide as a result of the first comparison process, and compare repeatedly the ID included in the signal emitted from the old RFID tag used in the first comparison process and the ID included in the signal emitted from another new RFID tag, which is different from the new RFID tag that emitted the signal used in the first comparison process, in response to the readout signal output continuously, for the another new RFID tag that is different for each comparison; and
a comparison result output unit configured to output a result of the comparison.

US Pat. No. 10,970,608

FINANCIAL SERVICES CARDS INCLUDING BRAILLE

United Services Automobil...

1. A card, comprising:a type of card;
a substrate;
multiple rows of braille-encoded characters formed on the substrate; and
a symbol being braille-encoded in at least one of the multiple rows indicating the type of card.

US Pat. No. 10,970,607

SECURE QR CODE USING NONLINEARITY OF SPATIAL FREQUENCY IN LIGHT

SHANGHAI JIAO TONG UNIVER...

1. A secure quick-response code (QrCode) communication method based on nonlinear spatial frequency, comprising:camera modeling: modeling based on the spatial frequency of the color filter array of the scanning device's camera;
QrCode encryption: using the color filter array (CFA) spatial frequency of the scanning device's camera and its modeling result, and the spatial frequency of the display device, to generate an encrypted picture of the QrCode on the display device;
QrCode decryption: the camera of the scanning device captures the encrypted QrCode shown on the display at a specified position and a specified angle, analyzes the target QrCode and recovers the QrCode,
wherein the modeling process specifically models the color filter array of the green filter layer,
wherein the modeling result is:
mcfa(x,y)=pcfa(?cfa(x,y))
pcfa(u)=0.5+0.5 cos(2?u)
?cfa(x,y)=((x,y)mod 2)/2
where, mcfa(x,y) is the green receiving information of the green filter layer on the camera image sensor at coordinates (x,y), pcfa(?) is the periodic function of the green filter layer in the CFA, and ?cfa (x,y) is the phase function of the green filter layer, and
wherein the QrCode decryption process specifically comprises:
Step B1: acquiring a plurality of consecutive pictures captured by the camera in the specified position and the specified angle to encrypt the QrCode in the display;
Step B2: selecting an unprocessed image and converting the RGB image to the HSV coordinates to maximize the saturation dimension;
Step B3: according to the positions of the three positioning marks, determining the size of the QrCode and dividing the QrCode for the converted picture to obtain a plurality of blocks;
Step B4: extracting a green square, for each pixel point, when the green intensity is greater than a predetermined threshold, it is judged to be white, and vice versa, it is judged to be black;
Step B5: for each square, determine the color of the square according to the distribution of white and black pixel points in the square;
Step B6: traversing all the black squares and mark all adjacent black squares to the same index value;
Step B7: determining whether all the blocks are marked or unmarked squares are all surrounded by the marked blocks, if yes, proceed to step B8, otherwise, return to step B2; and
Step B8: coloring according to the index value of each block, if two adjacent squares have the same index value, they are colored into the same color, and vice versa.

US Pat. No. 10,970,606

CONNECTOR HAVING COVER, BASE AND METAL MEMBER WITH A LOCKING PORTION, A COUPLING PORTION WITH A STANDING PORTION, AND A REINFORCING PORTION OVERLAPPING AN AREA OF TWICE A PREDETERMINED DISTANCE FROM THE STANDING PORTION

Japan Aviation Electronic...

1. A connector mountable on a circuit board, wherein:the connector comprises a cover, a base, a hinge mechanism, a terminal and a metal member;
the cover is attached to the base by the hinge mechanism;
the base has a front end and a rear end in a front-rear direction;
the hinge mechanism is closer to the front end of the base than to the rear end of the base in the front-rear direction;
the cover is provided with a locked portion;
the metal member is held by the base;
the metal member has a locking portion, a coupling portion and a reinforcing portion;
the locking portion has a first regulating portion and a second regulating portion;
the first regulating portion has a front end in the front-rear direction;
the second regulating portion extends downward in an up-down direction perpendicular to the front-rear direction from the front end of the first regulating portion;
the coupling portion integrally couples the first regulating portion and the reinforcing portion with each other;
the coupling portion and the second regulating portion are positioned away from each other by a predetermined distance in the front-rear direction;
the coupling portion has a standing portion;
the standing portion extends at least downward in the up-down direction from the first regulating portion;
the reinforcing portion is soldered on the circuit board when the connector is mounted on the circuit board;
in a plane perpendicular to the up-down direction, the reinforcing portion at least overlaps with an area within a range of twice the predetermined distance from the standing portion; and
the locked portion is configured to be pressed against the first regulating portion from below in the up-down direction.

US Pat. No. 10,970,605

ELECTRONIC APPARATUS AND METHOD OF OPERATING THE SAME

Samsung Electronics Co., ...

1. An electronic apparatus comprising:a display;
a memory configured to store at least one instruction; and
a processor configured to execute the at least one instruction stored in the memory,
wherein the processor is configured to execute the at least one instruction to cause the electronic apparatus to:
obtain a plurality of images,
extract deep features with respect to the plurality of images using a pre-trained feature extraction model that is based on general-purpose data,
classify the plurality of images into certain groups based on the extracted deep features and a pre-trained classification model that is based on general-purpose data,
display a result of the classification on the display,
determine, based on the result of the classification based on the extracted deep features, whether the feature extraction model and/or the classification model requires updating, and
when it is determined to update at least one of the feature extraction model and the classification model, re-train and update at least one of the feature extraction model and the classification model based on at least data of the user of the electronic device,
wherein the processor is configured to determine whether the feature extraction model and/or the classification model require updating based on a balance of the number of images included in the certain groups.

US Pat. No. 10,970,604

FUSION-BASED CLASSIFIER, CLASSIFICATION METHOD, AND CLASSIFICATION SYSTEM

Industrial Technology Res...

1. A fusion-based classifier, the classifier comprising:a storage medium, stores a plurality of modules; and
a processor, coupled to the storage medium, wherein the processor accesses and executes the plurality of modules, wherein the plurality of modules comprise:
a sub-classifier, generating a plurality of probability vectors according to input data, wherein each of the plurality of probability vectors comprises a plurality of elements respectively corresponding to a plurality of classes;
a fusion layer, selecting a first probability vector having a first class element corresponding to a first class-of-interest from the plurality of probability vectors according to the first class-of-interest, wherein the first class element has an extremum value between a plurality of first class elements corresponding to the plurality of probability vectors respectively, wherein the plurality of first class elements corresponding to the first class-of-interest; and
an output layer, determining a class of the input data according to the first probability vector.

US Pat. No. 10,970,603

OBJECT RECOGNITION AND DESCRIPTION USING MULTIMODAL RECURRENT NEURAL NETWORK

International Business Ma...

1. A method for image identification and classification, the method comprising:receiving, by a computer device, one or more images of a first object from at least two angles;
receiving, by the computing device, linguistic data associated with the first object, wherein the linguistic data of the first object describes the artist, art medium, age, color, symbol, pattern, function, and motif of the first object;
inputting, by the computing device, the one or more images of the first object into one or more first neural networks;
inputting, by the computing device, the linguistic data of the first object into one or more second neural networks;
combining, by the computing device, an output of the one or more first neural networks and the one or more second neural networks;
generating, by the computing device, an identification model based on the combined output of the one or more first neural networks and the one or more second neural networks, wherein the identification model generates a linguistic description for an unknown object;
receiving, by the computer device, at least one image of an unknown second object, wherein the second object is the unknown object (multiple images from different angles);
inputting, by the computer device, the at least one image of the unknown second object into the identification model to generate a linguistic description of the unknown second object;
analyzing, by the computer device, the at least one image of the unknown second object to identify different features of the unknown second object;
generating, by the computer device, a novel linguistic description identifying the unknown second object based on the identified different feature of the unknown second object, wherein the linguistic description includes a novel description of the unknown second object describing the unknown second object and the identified features of the unknown second object, wherein the generated linguistic description is based on a probability distribution of generating a word given previous linguistic data on the second neural networks and the one or more images on the first neural networks; and
displaying, by the computer device, the novel linguistic description identifying the unknown second object to a user.

US Pat. No. 10,970,602

SYSTEMS AND METHODS FOR CONVERTING VIDEO INFORMATION INTO ELECTRONIC OUTPUT FILES

Mythical, Inc., Sherman ...

1. A system configured for converting video information into electronic output files, the system comprising:one or more hardware processors configured by machine-readable instructions to:
obtain video information defining one or more videos, the one or more videos including a particular video, the particular video including video content having a progress length, the video content including visual content viewable as a function of progress through the progress length, wherein the visual content has a first electronic format, wherein the progress length ranges between 3 seconds and 1 minute when viewed at a frame rate between 20 frames per second and 60 frames per second, wherein the visual content is included within one or more images and/or one or more video frames of the particular video, wherein the video content has been captured by one or more capture devices;
select a subset of the visual content included in the video content of the particular video, wherein the subset is at least 20 times smaller than the visual content;
perform object detection to detect objects in the selected subset, wherein the object detection includes multiclass classification of a set of classes, wherein the set of classes includes a first class that represents living entities;
perform motion detection to detect a movement for one or more of the detected objects in the selected subset, responsive to the object detection detecting one or more detected objects; and
generate and store an electronic output file, responsive to the motion detection detecting the movement, wherein the electronic output file is in a second electronic format, wherein the second electronic format is a raster-graphics file format, and wherein the electronic output file includes one or more representations of the detected objects for which the movement has been detected by the motion detection.

US Pat. No. 10,970,601

INTERMEDIATE PROCESS STATE ESTIMATION METHOD

TOYOTA JIDOSHA KABUSHIKI ...

1. An intermediate process state estimation method for estimating, in time series, an intermediate process state in a deformation process optimal for molding from an initial state to a final state for a molding process that proceeds to the final state from the initial state through a plurality of intermediate molded objects, the method comprising:generating, by a first generator and a second generator, generation data based on a first label, a second label, and a common input noise, where the first and second labels correspond to steps included in a plurality of steps of the molding process, the first label corresponding to a certain step among the plurality of steps is input to the first generator and the second label corresponding to a different step from the certain step among the plurality of steps is input to the second generator;
determining whether an input data pair, randomly input to a discriminator, is a generation data pair or a training data pair so that the first and second generators and the discriminator learn in an adversarial manner from an accuracy of a discrimination result in the discriminator, where the generation data pair is generation data that corresponds to the first label that is generated by the first generator and the second label that is generated by the second generator, and the training data pair is data that a third label corresponding to training data that corresponds to the first label input to the first generator is linked to and data that a fourth label corresponding to training data that corresponds to the second label input to the second generator is linked to, and
inputting an input noise corresponding to a desired final state molded in the molding process that proceeds to the final state from the initial state through the plurality of intermediate molded objects and a label corresponding to a step included in the plurality of steps where it is desired that an intermediate process state be estimated to one of the respective learned generators to estimate, based on the generation data generated by the generator, the intermediate process state in the step where it is desired that the intermediate process state be estimated.

US Pat. No. 10,970,600

METHOD AND APPARATUS FOR TRAINING NEURAL NETWORK MODEL USED FOR IMAGE PROCESSING, AND STORAGE MEDIUM

TENCENT TECHNOLOGY (SHENZ...

1. A method for training a neural network model used for image processing, the method comprising:obtaining, by an electronic device comprising a memory storing instructions and a processor in communication with the memory, a plurality of video frames comprising an early video frame and a later video frame, wherein the early video frame is earlier in time than the later video frame;
inputting, by the electronic device, the plurality of video frames through a neural network model so that the neural network model outputs intermediate images corresponding to the plurality of video frames;
obtaining, by the electronic device, optical flow information between the early video frame and the later video frame;
modifying, by the electronic device according to the optical flow information, an intermediate image output from the neural network model and corresponding to the early video time frame to obtain an expected-intermediate image;
determining, by the electronic device, a time loss between an intermediate image output from the neural network model and corresponding to the later video frame and the expected-intermediate image;
determining, by the electronic device, a feature loss between each of the intermediate images corresponding to the plurality of video frames and a target feature image; and
training, by the electronic device, the neural network model according to the time loss and the feature loss, and returning to the act of obtaining a plurality of video frames to continue training until the neural network model satisfies a training finishing condition.

US Pat. No. 10,970,599

LEARNING COPY SPACE USING REGRESSION AND SEGMENTATION NEURAL NETWORKS

ADOBE INC., San Jose, CA...

1. A method for learning copy space of an image, the method comprising:applying, by a processor-based system, a regression convolutional neural network (CNN) to an image, the regression CNN to make a prediction of properties of a copy space, the properties including size and type, the prediction conditioned on a determination by the regression CNN of a copy space presence in the image;
applying, by the processor-based system, a segmentation CNN to the image, the segmentation CNN to generate one or more masks associated with locations of one or more of a manufactured copy space in the image, a natural copy space in the image, and a background region of the image;
searching a database of images for an image that includes a copy space associated with desired properties, the searching based on results from the regression CNN and the segmentation CNN; and
identifying at least one image that includes the copy space associated with the desired properties.

US Pat. No. 10,970,598

LEARNING METHOD AND LEARNING DEVICE FOR TRAINING AN OBJECT DETECTION NETWORK BY USING ATTENTION MAPS AND TESTING METHOD AND TESTING DEVICE USING THE SAME

Stradvision, Inc.

1. A method for training an object detection network by using one or more attention maps, comprising steps of:(a) an on-device learning device performing or supporting another device to perform (i) a process of inputting training images into a feature extraction network, to thereby allow the feature extraction network to perform one or more first convolution operations on the training images and thus to output one or more feature maps for training, (ii) a process of inputting the feature maps for training into an attention network, to thereby allow the attention network to output at least part of (ii-1) one or more attention maps for training corresponding to information on object densities for training of the feature maps for training and (ii-2) one or more prediction maps for training corresponding to the feature maps for training, and then (iii) a process of inputting the feature maps for training and the attention maps for training into a concatenation layer, to thereby allow the concatenation layer to concatenate the feature maps for training and the attention maps for training and thus to output one or more attentional feature maps for training;
(b) the on-device learning device performing or supporting another device to perform (i) a process of inputting the attentional feature maps for training into an RPN, to thereby allow the RPN to output one or more candidate boxes for training corresponding to one or more object candidates for training on the attentional feature maps for training, (ii) a process of inputting the candidate boxes for training into a binary convertor, to thereby allow the binary convertor to convert the candidate boxes for training to one or more binary maps for training, (iii) a process of inputting the attentional feature maps for training and the candidate boxes for training into an ROI pooling layer, to thereby allow the ROI pooling layer to perform one or more pooling operations on each of specific sections on the attentional feature maps for training, each of which corresponds to each of the candidate boxes for training, and thus to output one or more pooled feature maps for training, and (iv) a process of inputting the pooled feature maps for training into a detection network, to thereby allow the detection network to perform one or more learning operations on the pooled feature maps for training and thus to output each of object detection data for training each of which corresponds to each of the candidate boxes for training; and
(c) the on-device learning device performing or supporting another device to perform (i) a process of allowing an object loss layer to generate one or more object detection losses by referring to the object detection data for training and one or more object detection ground truths, to thereby train at least one of the feature extraction network and the detection network through one or more backpropagations using the object detection losses, (ii) a process of allowing a RPN loss layer to generate one or more RPN losses by referring to the candidate boxes for training and one or more RPN ground truths, to thereby train the RPN through one or more backpropagations using the RPN losses, and (iii) a process of allowing a cross-distilled loss layer to generate one or more cross-entropy losses by referring to the prediction maps for training and the binary maps for training, to thereby train the attention network through one or more backpropagations using the cross-entropy losses.

US Pat. No. 10,970,597

METHOD, APPARATUS, AND SYSTEM FOR PRIORITY RANKING OF SATELLITE IMAGES

HERE GLOBAL B.V., Eindho...

1. A method comprising:processing a plurality of images using a feature detector to determine a set of features on each image of the plurality of images;
determining a count of feature correspondences between each pair of images of the plurality of images based on the set of features of said each image;
computing a ranking of the plurality of images based on the count of features correspondences between said each pair of images; and
providing the ranking of the plurality of images as an output for selecting one or more images of the plurality of images for feature creation.

US Pat. No. 10,970,596

LOGISTIC REGRESSION GRADIENT CALCULATION METHOD AND APPARATUS

ALIBABA GROUP HOLDING LIM...

1. A logistic regression gradient calculation method performed by a computing machine, the method comprising:acquiring training data including X-row user data and Y-row click-through data corresponding to the X-row user data;
converting the X-row user data into X-column data;
segmenting the X-column data and a weight vector to form N X-column data segmentation blocks and N weight vector segmentation blocks, wherein N corresponds to a number of CPUs callable in the computing machine;
starting N threads to generate N sub-logistic regression gradients according to the N X-column data segmentation blocks, the N weight vector segmentation blocks, and the Y-row click-through data; and
splicing the N sub-logistic regression gradients to form a full logistic regression gradient.

US Pat. No. 10,970,595

METHODS AND SYSTEMS FOR DOCUMENT CLASSIFICATION USING MACHINE LEARNING

NETAPP, INC., Sunnyvale,...

1. A method, comprising:generating, by a processor, a plurality of topics from content of a plurality of electronic documents, each topic includes a plurality of words associated with the plurality of electronic documents;
determining, by the processor, a maximum distance between at least two words of each topic indicating compactness of each topic;
determining, by the processor, an average distance between word pairs of at least two topics of the plurality of topics indicating separation between the plurality of topics;
identifying, by the processor, a threshold score based on the maximum distance and the average distance;
utilizing, by the processor, the threshold score to select a subset of topics from the plurality of topics to represent the plurality of electronic documents;
automatically, generating by the processor, a tag for each subset topic, based on the tag's position within the subset topic; wherein each tag is an attribute of each subset topic;
storing, by the processor, the subset of topics with corresponding tags in a model data structure;
identifying, by the processor, a new topic and a new tag associated with a new electronic document; and
updating, by the processor, the model data structure based on the new topic and the new tag.

US Pat. No. 10,970,594

MAGNETIC RESONANCE MAPS FOR ANALYZING TISSUE

Tel HaShomer Medical Rese...

1. A method of magnetic resonance imaging (MRI), comprising:administering contrast agent to a subject;
acquiring a first MRI scan of a tissue, after said contrast administration;
acquiring a second MRI scan of said tissue, at least 20 minutes after said acquisition of said first MRI scan;
subtracting one of said MRI scans from another one of said MRI scans;
based on said subtraction, distinguish between a slow population, in which contrast clearance from the tissue is slower than contrast accumulation, and a fast population in which contrast clearance is faster than contrast accumulation; and
generating a displayed map identifying, in different colors, said fast population as an active tumor and said slow population as a tissue region other than an active tumor.

US Pat. No. 10,970,593

PRODUCT EFFECT ANALYSIS

International Business Ma...

1. A computer implemented method for determining effects of products on an individual, the method comprising:identifying, by one or more processors, a set of images of an individual and a product purchase history corresponding to the individual;
analyzing, by one or more processors, the set of images of the individual, the images being from a defined time window;
identifying, by one or more processors, one or more products within the product purchase history that correspond to use within the defined time window;
determining, by one or more processors, one or more changes in appearance of the individual based on the analyzed images;
determining, by one or more processors, correlations between the determined changes in appearance and the identified products used by the individual; and
notifying, by one or more processors, the individual of the determined correlations.

US Pat. No. 10,970,592

ADHERING SUBSTANCE DETECTION APPARATUS AND ADHERING SUBSTANCE DETECTION METHOD

DENSO TEN Limited, Hyogo...

1. An adhering substance detection apparatus comprising:a calculating unit that calculates a variation in a feature value related to luminance in past and current captured images captured by an image capturing device, based on the luminance of pixels included in the captured images;
a first detecting unit that detects a first region in which the variation calculated by the calculating unit falls within a predetermined threshold range, and in which the feature value in the current captured image falls within a predetermined threshold range;
a second detecting unit that detects a second region in which an irregularity in a distribution of the luminance of the pixels included in the captured image satisfies a predetermined irregularity condition; and
a generating unit that generates a sum region. that is a sum of the first region detected by the first detecting unit, and the second region detected by the second detecting unit, as an adhering substance region corresponding to an adhering substance adhering to the image capturing device.

US Pat. No. 10,970,591

AUTOMATED ACCURATE VIABLE SOLAR AREA DETERMINATION

OnSwitch LLC, San Franci...

1. A system for determining a viable area for solar radiation, comprising:an interface configured to:
receive location information; and
a processor configured to:
determine imagery data based on the location information;
determine a point cloud using the imagery data, wherein the determining of the point cloud comprises to:
determine points corresponding with a plurality of images of the imagery data;
determine coordinates associated with the points; and
project the points onto images of the earth to obtain a set of three dimensionally specified points;
determine ground points, vegetation points, and structure points using the point cloud, wherein determining the one or more features using the point cloud comprises to:
classify ground points based on a determination of points in a plane;
classify vegetation points based on a signature associated with points in an infrared image; and
classify structure points based on a determination of points that are above a ground plane;
determine a viable area using the ground points, vegetation points, and structure points, comprises to:
extract corners of a structure of the structure points, vegetation points, and structure points;
determine adjacent points based on the corners, wherein the adjacent points have the same orientation or tilt;
determine the viable area based on the adjacent points; and
determine solar power configuration based on the viable area.

US Pat. No. 10,970,590

IMAGE-BASED WELLSITE EQUIPMENT HEALTH MONITORING

Schlumberger Technology C...

1. An apparatus comprising:a processing system comprising a processor and a memory including computer program code, wherein the processing system is operable to:
identify a unique oilfield equipment unit of an oilfield wellsite by analyzing a visible image based on business rules relating to a plan for setting up and configuring a plurality of oilfield equipment units at the oilfield wellsite;
determine a numbering scheme for the plurality of oilfield equipment units based on the unique oilfield equipment unit and based on the business rules;
overlay the unique oilfield equipment unit of the visible image on a corresponding portion of a thermal image; and
generate an alert when a temperature indicated by the corresponding portion of the thermal image is outside of an operational temperature range for the unique oilfield equipment unit.

US Pat. No. 10,970,589

DOMAIN SEPARATION NEURAL NETWORKS

Google LLC, Mountain Vie...

1. A system comprising:a shared encoder neural network implemented by one or more computers, wherein the shared encoder neural network is configured to:
receive an input image from a target domain; and
process the input image to generate a shared feature representation of features of the input image that are shared between images from the target domain and images from a source domain different from the target domain, wherein the shared encoder neural network has been trained to generate a shared feature representation for the input image from the target domain that, when combined with a private feature representation for the same input image from the target domain generated by a private target encoder neural network that is specific to the target domain, can be used to accurately reconstruct the input image by a shared decoder neural network; and
a classifier neural network implemented by the one or more computers, wherein the classifier neural network is configured to:
receive the shared feature representation; and
process the shared feature representation to generate a network output for the input image that characterizes the input image.

US Pat. No. 10,970,588

RECURRENT DEEP NEURAL NETWORK SYSTEM FOR DETECTING OVERLAYS IN IMAGES

Gracenote, Inc., Emeryvi...

1. A method comprising:a processor applying a feature map network to an image to create a feature map, wherein the feature map comprises a grid of vectors characterizing at least one feature in the image;
the processor applying a probability map network to the feature map to create a probability map assigning a probability to the at least one feature in the image, wherein the assigned probability corresponds to a likelihood that the at least one feature is an overlay;
the processor determining that the assigned probability exceeds a threshold probability value; and
responsive to the processor determining that the assigned probability exceeds the threshold probability value, the processor (i) determining compliance with an advertising requirement, (ii) performing optical character recognition (OCR) on the at least one feature, (iii) determining a channel associated with the at least one feature, (iv) determining a show associated with the at least one feature, (v) determining a hyperlink associated with the at least one feature, or (vi) accessing a hyperlinked object associated with the at least one feature.

US Pat. No. 10,970,587

ELECTRONIC DEVICE FOR NOTIFYING OF UPDATE OF IMAGE SIGNAL PROCESSING AND METHOD FOR OPERATING THE SAME

Samsung Electronics Co., ...

1. An electronic device comprising:a display;
a memory;
a communication module comprising communication circuitry; and
a processor,
wherein the processor is configured to:
receive a second image corresponding to a first image stored in the memory from an external electronic device through the communication module, the second image being an image in which a raw image corresponding to the first image is corrected,
identify information associated with at least one stat used by the external electronic device to correct the second image,
identify an amount of change of the at least one stat of a first plurality of images including the first image stored in the memory and a second plurality of images including the second image corresponding to the first plurality of images received from the external electronic device,
select the first image having a highest amount of change of the stat with the second plurality of images from among the first plurality of images,
control the display to display the first image on a first segment of the display and to display the second image on a second segment of the display, and
identify whether to store the second image in the memory, based on an input for the first image or the second image.

US Pat. No. 10,970,586

SYSTEMS AND METHODS OF 3D SCENE SEGMENTATION AND MATCHING FOR ROBOTIC OPERATIONS

GENERAL ELECTRIC COMPANY,...

1. A system, comprising:a memory storing executable program instructions therein; and
a processor in communication with the memory, the processor operative to execute the program instructions to:
receive image data comprising a set of images of a physical asset;
receive data models of a plurality of assets, wherein each of the data models includes a semantic description of a respective asset, and wherein the semantic description includes at least one operation associated with the respective asset;
determine a match between the physical asset represented in the image data and a corresponding data model of the data models;
generate, upon determining the match between the physical asset and the corresponding data model, an operation plan based on a corresponding operation included in the corresponding data model;
execute the operation plan via a robot configured to interact with the physical asset;
quantify a result of execution of the operation plan by the robot;
update the corresponding operation included in the corresponding data model based on the result to generate an updated operation associated with the corresponding data model and different than the corresponding operation; and
store the updated operation in an operation library.

US Pat. No. 10,970,585

ADHERING SUBSTANCE DETECTION APPARATUS AND ADHERING SUBSTANCE DETECTION METHOD

DENSO TEN Limited, Kobe ...

20. An adhering substance detection method comprising:for each cell of a plurality of cells of a captured image, each of the cells being composed of a predetermined number of pixels in the captured image, (i) calculating an edge feature value that is based on edge vectors calculated for the pixels and (ii) classifying an edge orientation, which is included in the calculated edge feature value, to an angle class using two types of angle classifications; and
determining a condition of an adhering substance adhering to an image capturing apparatus that captured the captured image, based on a transition count representing a number of transitions of the angle class within a unit region, which is a predetermined region composed of a predetermined number of the cells.

US Pat. No. 10,970,584

SYSTEMS, DEVICES, AND METHODS FOR EVALUATING READINGS OF GAUGE DIALS

Intelligent Security Syst...

1. A method comprising:creating, by a processor, a patrol tour for a camera such that the patrol tour includes a plurality of points associated with a plurality of identifiers;
scheduling, by the processor, the patrol tour such that the camera can capture a plurality of images at the points during the patrol tour and surveil an area when not capturing the images during the patrol tour;
programming, by the processor, the camera such that the images are (a) associated with the identifiers during the patrol tour, a plurality of date stamps in accordance with when the images are captured during the patrol tour, and a plurality of time stamps in accordance with when the images are captured during the patrol tour, and (b) sent to a server with the identifiers, the date stamps, and the time stamps during the patrol tour, wherein the server is remote from the camera during the patrol tour;
causing, by the processor, the server to:
receive the images with the identifiers, the date stamps, and the time stamps during the patrol tour, wherein the images include a plurality of content;
identify the content within the images during the patrol tour; and
populate a data structure with the content during the patrol tour based on the identifiers, the date stamps, and the time stamps.

US Pat. No. 10,970,583

INFORMATION PROCESSING APPARATUS, AND NON-TRANSITORY COMPUTER READABLE MEDIUM

FUJI XEROX CO., LTD., To...

1. An information processing apparatus comprising:a processor, configured to:
generate and display on a display a drawing that expresses a meaning of target data; and
extract partial data that is at least part of the target data used to express a specific portion, which is in part or in whole of the drawing, in response to the specific portion being specified,
wherein the processor is configured to display the drawing including an expression of additional information that is not included in the target data, and not extract the additional information as the partial data,
wherein if the drawing generated in accordance with a remaining portion of the target data with the partial data removed therefrom is displayed after the drawing including the expression of the additional information is displayed, the processor is configured to display a screen prompting a user to edit the additional information on the display.

US Pat. No. 10,970,582

INFORMATION PROCESSING METHOD, INFORMATION PROCESSING DEVICE, AND RECORDING MEDIUM

Panasonic Intellectual Pr...

1. An information processing method performed by a computer, the information processing method comprising:obtaining a first image including a noise region;
obtaining noise region estimation information indicating the noise region which is estimated, the noise region estimation information being output from a first converter as a result of the first image being input to the first converter;
obtaining a second image on which noise region removal processing has been performed, the second image being output from a second converter as a result of the noise region estimation information and the first image being input to the second converter;
obtaining a third image which does not include the noise region, the third image including a scene identical to or corresponding to a scene included in the first image;
generating a fourth image by using the noise region estimation information and the third image, the fourth image including the noise region which is estimated;
training the first converter by using machine learning in which the first image is reference data and the fourth image is conversion data; and
training the second converter by using machine learning in which the third image is reference data and the second image is conversion data.

US Pat. No. 10,970,581

ELECTRONIC DEVICE

KYOCERA Document Solution...

1. An electronic device comprising:a reading section configured to read an image formed on a document to generate first image information;
an extraction section configured to extract a specific area from the image based on the first image information;
a recognition section configured to execute character recognition processing on information of a character string image included in the specific area to generate text information corresponding to the information of the character string image;
a search section configured to search for a webpage containing information relating to a meaning of a text indicated by the text information;
an attachment section configured to attach link information of the webpage to the information of the character string image to generate second image information; and
storage configured to store the second image information therein, wherein
the specific area is an area with a specific mark applied thereto,
the electronic device further comprises a selection section configured to select a first specific number of webpages,
the search section searches a first search site for the webpages and searches a second search site for the webpages, the second search site differing from the first search site,
the selection section selects the first specific number of webpages based on a result of search on the first search site and a result of search on the second search site,
the attachment section attaches, as a reference, link information of each of the first specific number of webpages to the information of the character string image,
the selection section selects the first specific number of webpages based on a second specific number of webpages among webpages found in the search on the first search site in descending order from a top and the second specific number of webpages among webpages found in the search on the second search site in descending order from a top,
the second specific number is larger than the first specific number, and
the first specific number of webpages is included in the second specific number of webpages among webpages found in the search on the first search site in the descending order from the top and the second specific number of webpages among webpages found in the search on the second search site in the descending order from the top.

US Pat. No. 10,970,580

INFORMATION PROCESSING APPARATUS, INFORMATION PROCESSING METHOD, AND STORAGE MEDIUM

Canon Kabushiki Kaisha, ...

1. An information processing apparatus having a blotting out function of an electronic document, the information processing apparatus comprising:a setting unit configured to set an area to which blotting out processing is applied, wherein the area is specified by a user; and
a blotting out processing unit configured to fill the inside of the area set by the setting unit for a character image object and to delete text information within the area set by the setting unit for a text object, of objects included in the electronic document, wherein
in a case where the electronic document is an electronic document in a format searchable for a character string and where a character string obtained by performing optical character recognition (OCR) processing for a character image object existing in the area and a character string of an invisible text object existing in the area do not match with each other, the setting unit performs re-setting of an area to which blotting out processing is applied for at least one of the character image object and the invisible text object so that both the character strings match with each other.

US Pat. No. 10,970,579

IMAGE PROCESSING APPARATUS FOR PLACING A CHARACTER RECOGNITION TARGET REGION AT A POSITION OF A PREDETERMINED REGION IN AN IMAGE CONFORMING TO A PREDETERMINED FORMAT

PFU LIMITED, Kahoku (JP)...

1. An image processing apparatus comprising:a storage device to store, in advance, positions of a plurality of predetermined machine readable zones (MRZs) in an image of a document having a predetermined format;
a processor to
acquire an input image of a document including a plurality of character recognition target regions, respectively corresponding to the plurality of stored predetermined MRZs and the predetermined format of the document,
generate a rotated binary image by binarizing a rotated image acquired by rotating the input image,
calculate a black pixel count on each of a plurality of lines extending in a predetermined direction in the rotated binary image, for each line extending in the predetermined direction,
detect the character recognition target region, based on the black pixel count calculated for each of the plurality of lines,
cut out a plurality of regions respectively corresponding to the plurality of character recognition target regions from the input image or an image generated from the input image,
place the cut out plurality of regions at positions that are respectively corresponding to the stored positions of the plurality of predetermined MRZs corresponding to the predetermined format of the document,
generate a corrected image in which the cut out plurality of regions are placed at the positions respectively corresponding to stored positions, and
detect a character from the corrected image; and
an output device to output information related to the detected, wherein the processor.

US Pat. No. 10,970,578

SYSTEM AND METHOD FOR EXTRACTING INFORMATION FROM A NON-PLANAR SURFACE

JOHNSON CONTROLS FIRE PRO...

1. A method of extracting data from an image, comprising:receiving images that capture a non-planar surface from different respective angles, wherein each of the images captures a section of an area on the non-planar surface, the area including textual or graphical information;
identifying low image quality sections corresponding to curved edges of the non-planar surface in the images;
cropping the images to remove the low image quality sections;
stitching together the images to obtain a stitched image that captures an entirety of the area on the non-planar surface;
processing the stitched image to extract the textual or graphical information from the area on the non-planar surface;
associating the textual or graphical information with one or more information fields in a report template; and
generating a report by populating the one or more information fields of the report template with corresponding associated information within the textual or graphical information.

US Pat. No. 10,970,577

MACHINE LEARNED SINGLE IMAGE ICON IDENTIFICATION

Snap Inc., Santa Monica,...

1. A method comprising:receiving, by one or more processors, an image including a graphical icon;
identifying, by the one or more processors, a set of proposed regions of the image, at least one proposed region of the set of proposed regions containing the graphical icon;
extracting a set of semantic features for each proposed region of the set of proposed regions;
based on the set of semantic features of the set of proposed regions, identifying, by the one or more processors, a set of proposed icons corresponding to the graphical icon included in the image; and
determining a match between the graphical icon and at least one proposed icon of the set of proposed icons, the determining of the match comprising performing a pairwise comparison of visually distinct aspects of the graphical icon, included in the image, and each proposed icon of the set of proposed icons that are identified based on the set of semantic features.

US Pat. No. 10,970,576

METHODS AND SYSTEMS FOR CAPTURING IMAGE DATA

DAON HOLDINGS LIMITED, G...

1. A method for capturing image data of a user comprising the steps of:displaying, by a computing device, a perspective view of a virtual computing device;
changing, by the computing device, the perspective view of the virtual computing device while the computing device moves;
ceasing, by the computing device, to display the virtual computing device to provide notification that the computing device is in an acceptable position for capturing image data of a user; and
capturing, by a camera included in the computing device, facial image data of the user.

US Pat. No. 10,970,575

METHODS, APPARATUS AND SYSTEMS FOR BIOMETRIC PROCESSES

Cirrus Logic, Inc., Aust...

1. A method operable in a biometric authentication system, the method comprising:initiating generation of an acoustic stimulus for application to a user's ear;
determining a quality measure of a response signal to the acoustic stimulus; andresponsive to determining that the quality measure is inadequate for performing a biometric process, modifying one or more properties of the acoustic stimulus to improve a signal to noise ratio, SNR, of the response signal;wherein determining that the quality measure is inadequate for performing a biometric process comprises comparing the quality measure with a target quality measure; and
wherein responsive to determining that a difference between the determined quality measure and a target quality measure is positive and greater than a threshold gain metric, reducing the gain of the acoustic stimulus and responsive to determining that the difference between the determined quality measure and the target quality measure is negative and greater than a threshold gain metric, increasing the gain of the acoustic stimulus.

US Pat. No. 10,970,574

SPOOF DETECTION USING DUAL-BAND NEAR-INFRARED (NIR) IMAGING

Advanced New Technologies...

1. A method for preventing access to a secure system based on determining that a subject is an alternative representation of a live person, the method comprising:illuminating a first subject with electromagnetic radiation in a first wavelength range;
capturing a first image of the first subject under illumination of the first wavelength range;
illuminating the first subject with electromagnetic radiation in a second wavelength range, wherein a difference between (i) absorption of the second wavelength range by live human tissue, and (ii) absorption of the first wavelength range by live human tissue satisfies a threshold condition;
capturing a second image of the first subject under illumination of the second wavelength range;
calculating pixel-wise differences in intensity values between corresponding pixels of the first image and the second image;
calculating an aggregated measure of the pixel-wise differences in intensity values;
calculating a difference between the aggregated measure of the pixel-wise differences and a target value; and
identifying, based at least on the difference between the aggregated measure and the target value, the first subject as an alternative representation of a live person; and
preventing access to the secure system responsive to identifying the first subject as an alternative representation of a live person.

US Pat. No. 10,970,573

METHOD AND SYSTEM FOR FREE TEXT KEYSTROKE BIOMETRIC AUTHENTICATION

1. A user authentication method for authenticating a user using a keystroke dynamics, the method comprising:receiving a user keystroke input implemented by the user typing free text;
separating a sequence of pressed keys based on the user keystroke input into a sequence of first bigrams having first bigram names simultaneously with the user typing free text;
collecting a first timing information for each first bigram of the sequence of first bigrams;
extracting a feature vector for each first bigram of the sequence of first bigrams based on the first timing information;
separating extracted feature vectors for each first bigram of the sequence of first bigrams into subsets according to the first bigram names;
estimating a Gaussian Mixture Model (GMM) user model using the subsets of feature vectors for each first bigram of the sequence of first bigrams using a Universal Background Model (UBM) as a predefined GMM; and
providing real time user authentication using the estimated GMM user model for each first bigram of the sequence of first bigrams and bigram feature vectors generated based on a current user keystroke input implemented by the user.

US Pat. No. 10,970,572

DRIVER CONDITION DETECTION SYSTEM

Toyota Jidosha Kabushiki ...

1. A driver condition detection system comprising:a driver monitor camera capturing a face of a driver of a vehicle and generating a facial image of the driver, and
a driver condition detection part configured to detect a condition of the driver based on the facial image, wherein:
when a part of face parts of the driver is hidden in the facial image, the driver condition detection part is configured to detect a condition of the driver based on each of face parts of the driver not hidden in the facial image, the face parts of the driver being a mouth, nose, right eye, and left eye of the driver,
the condition of the driver is a facial direction of the driver,
when a part of the face parts of the driver is hidden in the facial image, the driver condition detection part is configured to detect the facial direction of the driver based on the face parts of the driver not hidden in the facial image and the facial image before a part of the face parts of the driver is hidden, and
the driver condition detection part is configured to judge that the facial direction of the driver has not changed from before a part of the face parts of the driver was hidden when an amount of change of a position and size of the face parts of the driver not hidden in the facial image is a threshold value or less, and not detect the facial direction of the driver when the amount of change of at least one of the position and size is larger than the threshold value.

US Pat. No. 10,970,571

VEHICLE CONTROL METHOD AND SYSTEM, VEHICLE-MOUNTED INTELLIGENT SYSTEM, ELECTRONIC DEVICE, AND MEDIUM

SHANGHAI SENSETIME INTELL...

1. A vehicle control method, performed by a vehicle device, comprising:acquiring, by a photographing component provided on a vehicle, a face image of a user currently requesting to use the vehicle, wherein the operation of using the vehicle comprises at least one of reserving use of a vehicle, driving a vehicle, riding a vehicle, cleaning a vehicle, maintaining a vehicle, repairing a vehicle, refueling a vehicle, or charging a vehicle;
acquiring a result of a feature matching between the face image and at least one pre-stored face image in a data set of the vehicle, wherein the data set stores one or more pre-stored face images of at least one pre-recorded user allowed to use the vehicle, and the data set stores the pre-stored face image of at least one user having reserved a ride; and
in response to that the result of the feature matching indicates that the feature matching is successful, controlling actions of the vehicle to allow the user to use the vehicle, otherwise, controlling the actions of the vehicle to refuse the user to use the vehicle, wherein the operation of controlling actions of the vehicle to allow the user to use the vehicle comprises at least one of: controlling opening of vehicle doors, release of driving control rights of the vehicle, controlling opening of a vehicle lock, controlling opening of a fuel filler of the vehicle or performing control to allow a charging device to connect to a battery of the vehicle.

US Pat. No. 10,970,570

DEVICE AND A METHOD FOR CLUSTERING LIGHT SPOTS

Aptiv Technologies Limite...

1. A device for clustering light spots,wherein the light spots are spots of headlights and tail lights of vehicles;
the device is configured to:
receive an image captured by a camera;
detect light spots in the image;
calculate a distance measure for each of a plurality of pairs of light spots detected in the image based on two or more features of the two light spots of the respective pair, wherein the distance measure quantifies a dissimilarity of features between each of the plurality of pairs of light spots; and
group the detected light spots to form clusters based on the calculated distance measures,
wherein two light spots form a new cluster if their distance measure is below a predetermined threshold T2, and
a third light spot is assigned to the cluster if a distance measure to at least one of the two light spots of the cluster is below the predetermined threshold T2.

US Pat. No. 10,970,569

SYSTEMS AND METHODS FOR MONITORING TRAFFIC LIGHTS USING IMAGING SENSORS OF VEHICLES

TOYOTA MOTOR NORTH AMERIC...

1. A traffic light monitoring system comprising:a vehicle configured to:
identify a traffic light based on images captured by an imaging sensor of the vehicle;
determine a first transition time when a first light of a traffic light turns off and a second light of the traffic light turns on based on the images;
determine a second transition time when the second light of the traffic light turns off and a third light of the traffic light turns on based on the images; and
transmit information about the first transition time and the second transition time to an edge computing device; and
the edge computing device configured to:
calculate a period for the second light being on based on the first transition time and the second transition time;
receive another transition time when the first light turns off and the second light turns on from another vehicle, the another transition time being after the first transition time; and
broadcast a notification of an upcoming change of the second light based on the another transition time, the period, and a current time.

US Pat. No. 10,970,568

VEHICULAR VISION SYSTEM WITH OBJECT DETECTION

MAGNA ELECTRONICS INC., ...

1. A vehicular vision system, the vehicular vision system comprising:a camera disposed at a vehicle and viewing exterior of the vehicle, the camera capturing frames of image data;
a control having an image processor operable to process multiple frames of image data captured by the camera;
wherein the control receives vehicle motion data indicative of motion of the vehicle when the vehicle is moving;
wherein, when the vehicle is moving, the control, responsive at least in part to image processing of multiple frames of captured image data, detects an object present exterior of the vehicle; and
wherein the control, responsive at least in part to the received vehicle motion data and image processing of multiple frames of captured image data, determines motion of the detected object relative to the moving vehicle by (i) determining corresponding object points in at least two frames of captured image data and (ii) estimating object motion trajectory of the detected object based at least in part on the determined corresponding object points.

US Pat. No. 10,970,567

METHOD AND SYSTEM FOR SYNTHESIZING A LANE IMAGE

Catch Technology, Inc., ...

1. An image synthesizing system of a vehicle driven on a road having a dashed line having a plurality of dashes and a plurality of spaces, each of which is located between every adjacent two dashes, the image synthesizing system comprising:a video image capture device obtaining a plurality of image frames having a plurality of fractional images of the dashes at a frame rate f; and
an image mapping module configured to obtain the frame rate f from the video image capture device and a real-time velocity of the vehicle v, and including:
an image register configured to store the plurality of image frames;
an image mapping calculator determining a quantity N for an image mapping based on a dash length L and a distance S between two dashes of the dashed line;
a frame interval calculator determining a frame interval m based on the real-time velocity v, the dash length L, the distance S and the frame rate f; and
an image composer fetching at least N image frames from the plurality of image frames according to the frame interval m, and synthesizing the at least N image frames to obtain a virtual solid line based on fractional images of the dashes in the fetched at least N image frames.

US Pat. No. 10,970,566

LANE LINE DETECTION METHOD AND APPARATUS

BOE TECHNOLOGY GROUP CO.,...

1. A lane line detection method applicable for an in-vehicle device and comprising:determining a region of interest in an image to be detected;
extracting lane line pixel features in the region of interest;
combining similar lane line pixel features to generate superpixels corresponding to the combined lane line pixel features; and
performing a clustering and fitting process for the superpixels to obtain a target lane line,
wherein the step of determining the region of interest in the image to detected comprises:
setting a lane line procxing region around the in-vehicle device;
determining coordinate values of a midpoint on each of boundary lines of the lane line processing region in a real coordinate system in which the in-vehicle device is located;
converting each of the coordinate values into a corresponding image coordinate value in an image coordinate system corresponding to the image to be detected; and
determining the region of interest in the image to be detected according to the respective image coordinate values.

US Pat. No. 10,970,565

METHOD AND APPARATUS WITH ROAD LINE DETECTION

Samsung Electronics Co., ...

1. A processor-implemented method of detecting a road line, comprising:segmenting a captured driving image data into a plurality of segmentation areas;
determining a candidate vanishing-point area corresponding to a segmentation area of the segmentation areas;
extracting at least one straight road line from the segmentation area;
detecting a partial line corresponding to the segmentation area based on whether the at least one straight road line meets the candidate vanishing-point area; and
detecting the road line of the driving image data by connecting partial lines corresponding to the segmentation areas,
wherein the determining of the candidate vanishing-point area comprises:
verifying whether the segmentation area is a lowermost segmentation area of the driving image data; and
extracting, when the segmentation area is not the lowermost segmentation area, a candidate vanishing-point area corresponding to the segmentation area based on a partial vanishing point determined in a preceding segmentation area adjacent to the segmentation area.

US Pat. No. 10,970,564

SYSTEM AND METHOD FOR INSTANCE-LEVEL LANE DETECTION FOR AUTONOMOUS VEHICLE CONTROL

TUSIMPLE, INC., San Dieg...

1. A system comprising:a data processor; and
a memory for storing an autonomous vehicle instance-level lane detection system, executable by the data processor, the autonomous vehicle instance-level lane detection system being configured to:
receive training image data from a training image data collection system, the training image data collection system comprising sensors installed in a moving test vehicle navigated through real-world traffic scenarios, the training image data comprising data collected from real-world traffic scenarios;
obtain ground truth data corresponding to the training image data, the ground truth data corresponding to the data collected from real-world traffic scenarios;
perform a training phase to train a plurality of tasks each associated with different features of the training image data, at least one task of the plurality of tasks corresponding to a specific feature of the training image data, the plurality of tasks configured to execute concurrently, the training phase comprising extracting roadway lane marking features from the training image data, associating similar extracted features with a corresponding task of the plurality of tasks, associating different extracted features with different other tasks of the plurality of tasks, causing the plurality of tasks to generate task-specific predictions of feature characteristics based on the training image data, determining a bias between the task-specific prediction for each task and corresponding task-specific ground truth data, and adjusting parameters of each of the plurality of tasks to cause the bias to meet a pre-defined confidence level;
receive image data from an image data collection system associated with an autonomous vehicle; and
perform an operational phase comprising extracting roadway lane marking features from the image data, causing the plurality of trained tasks to execute concurrently to generate instance-level lane detection results based on the image data, and providing the instance-level lane detection results to an autonomous vehicle subsystem of the autonomous vehicle.

US Pat. No. 10,970,563

SYSTEM AND METHOD FOR REMAPPING SURFACE AREAS OF A VEHICLE ENVIRONMENT

Honda Motor Co., Ltd., T...

1. A computer-implemented method for remapping surface areas of a vehicle environment, comprising:receiving a plurality of images of the surface areas of the vehicle environment from a computing device, wherein the surface areas of the vehicle environment includes an internal environment of a vehicle and an external environment surrounding the vehicle;
evaluating the plurality of images to determine optical flow values associated with at least one pixel of at least two images of the plurality of images;
determining at least one surface area that includes the external environment of the vehicle, wherein the optical flow values associated with the at least one pixel of the at least two images are tracked and plotted to determine the surface area that includes the external environment of the vehicle; and
remapping the surface areas by filtering the at least one surface area included within the at least one pixel that includes the external environment and rendering at least one surface area graphical augmentation on at least one remapped surface area of the vehicle environment.

US Pat. No. 10,970,562

DISPLAY APPARATUS

PANASONIC INTELLECTUAL PR...

1. A display apparatus for a vehicle comprising:a first camera configured to capture first images of first ambient view at a first out portion of the vehicle;
a second camera configured to capture second images of second ambient view at a second out portion of the vehicle;
a third camera configured to capture third images of third ambient view at a third out portion of the vehicle, a first distance between the first out portion and the third out portion being longer than a second distance between the first out portion and the second out portion, the first distance being longer than a third distance between the second out portion and the third out portion;
a sensor having a detection area at the second out portion of the vehicle, and being capable of detect a three-dimensional object within the detection area; and
a display configured to display at least a first one image and a second one image,
wherein:
when the sensor does not detect a three-dimensional object at the second out portion of the vehicle, the display displays the first one image having at least a first region and a second region, the first region being adjacent to the second region in the first one image, the first region corresponding to one of the first images, the second region corresponding to one of the third images,
when the sensor detects a three-dimensional object at the second out portion of the vehicle, the display displays the second one image having at least a third region, a fourth region, and a fifth region, the third region being adjacent to the fourth region in the second one image, the fourth region being adjacent to the fifth region in the second one image, the third region corresponding to one of the first images, the fourth region corresponding to one of the second images, the fifth region corresponding to one of the third images, and
the fourth region of the second one image includes a whole image of the detection area of the sensor, the whole image including at least a part of the three-dimensional object.

US Pat. No. 10,970,561

DETECTING ROADWAY OBJECTS IN REAL-TIME IMAGES

1. A method for identifying roadway objects, the method comprising:receiving a real-time image of a road from a camera sensor communicatively coupled to an onboard computer of a vehicle;
dividing the real-time image into superpixels;
merging the superpixels to form superpixel regions based on similarity of the superpixels;
drawing initial bounding boxes where each initial bounding box surrounds one of the superpixel regions;
identifying a road region and a central horizon line in the real-time image;
determining a proximity of each of the initial bounding boxes to the road region and the central horizon line in the real-time image;
pruning the initial bounding boxes with a centroid value located above the central horizon line to reduce the initial bounding boxes to remaining bounding boxes;
estimating a size of each of the remaining bounding boxes in the real-time image; and
identifying a set of objects in the road region based on the proximity of the remaining bounding boxes to the road region, the proximity of the remaining bounding boxes to the central horizon line, and the size of each of the remaining bounding boxes.

US Pat. No. 10,970,560

SYSTEMS AND METHODS TO TRIGGER PRESENTATION OF IN-VEHICLE CONTENT

Disney Enterprises, Inc.,...

1. A system configured to trigger presentation of in-vehicle content based on detected physical objects, the system being coupled to a vehicle, the system comprising:one or more physical processors configured by machine readable instructions to:
obtain output signals conveying detection information, the detection information facilitating detection and identification of physical objects near and/or within an interior of the vehicle;
detect, based on the output signals, presence of one or more physical objects near and/or within the interior of the vehicle, including detecting presence of a first physical object near and/or within the interior of the vehicle;
identify, based on the output signals, the one or more physical objects detected as being present near and/or within the interior of the vehicle, including identifying the first physical object;
determine content that is specific to the identity of the one or more physical objects detected as being present near and/or within the interior of the vehicle, including determining first content that is specific to the identity of the first physical object;
control one or more vehicle components coupled to the vehicle to provide the content that is specific to the identity of the one or more physical objects, including controlling a first vehicle component to provide the first content; and
responsive to detecting a combination of two or more physical objects as being concurrently present near and/or within the interior of the vehicle, identify the combination of two or more physical objects, determine content that is specific to the combination of the identities of two or more physical objects, and control one or more vehicle components coupled to the vehicle to provide the content that is specific to the identity of the combination of two or more physical objects, such that:
responsive to detecting presence of a second physical object near and/or within the interior of the vehicle concurrently with the first physical object:
identify the second physical object;
determine second content that is specific to the combination of the identities of the first physical object and the second physical object, the second content including a combination of content specific to the identities of both the first physical object and the second physical object; and
control a first vehicle component to provide the second content.

US Pat. No. 10,970,559

PEOPLE FLOW ESTIMATION DEVICE, PEOPLE FLOW ESTIMATION METHOD, AND RECORDING MEDIUM

NEC CORPORATION, Tokyo (...

1. A people flow estimation system comprising:at least one memory storing a computer program; and
at least one processor executing the computer program to perform:
estimating, for each partial area of an input image, a number of persons within the partial area;
calculating a number-of-people correction coefficient, for each partial area of the input image, based on the number of persons within the partial area, and a correction value with which the number of persons is corrected;
generating a weight of each particle representing a probability of a state of each particle, based on a particle probability hypothesis density (PHD) filter and the number-of-people correction coefficient; and
outputting, as a flow of people, a state of particles resampled based on the weight.

US Pat. No. 10,970,558

PEOPLE FLOW ESTIMATION DEVICE, PEOPLE FLOW ESTIMATION METHOD, AND RECORDING MEDIUM

NEC CORPORATION, Tokyo (...

1. A people flow estimation system comprising:at least one memory storing a computer program; and
at least one processor executing the computer program to perform:
estimating, for each partial area of an input image, a number of persons within the partial area;
calculating a number-of-people correction coefficient, for each partial area of the input image, based on the number of persons within the partial area, and a correction value with which the number of persons is corrected;
generating a weight of each particle representing a probability of a state of each particle, based on a particle probability hypothesis density (PHD) filter and the number-of-people correction coefficient;
merging particles on the basis of positions of the particles, moving velocities of the particles and weights of the particles; and
displaying the merged particles as an object on a display device.

US Pat. No. 10,970,557

POSTURE DETERMINATION METHOD, ELECTRONIC SYSTEM AND NON-TRANSITORY COMPUTER-READABLE RECORDING MEDIUM

Wistron Corporation, New...

1. A posture determination method, applicable to an electronic system comprising an image capturing device, wherein the image capturing device is set up corresponding to a target, and the posture determination method comprises:acquiring a plurality of consecutive images captured by the image capturing device;
performing a subtraction operation on two images selected from the consecutive images to generate an image difference result so as to determine whether an image content of the consecutive images has a movement;
determining whether the image content of the consecutive images is static after the movement according to the image difference result; and
determining a posture of the target in response to that the image content is static after the movement, wherein the target is a head contour,
wherein the step of determining the posture of the target in response to that the image content is static after the movement comprises:
calculating a degree of similarity between a template image of the target and a reference image of the target, wherein the template image of the target is a target image corresponding to the image content in the movement, and the reference image of the target is an initial target image captured by the image capturing device or a normalized target image.

US Pat. No. 10,970,556

SMART SURVEILLANCE CAMERA SYSTEMS AND METHODS

FLIR SYSTEMS, INC., Wils...

13. A method comprising:capturing, at a focal plane array (FPA) of a thermal imager, thermal images of a scene;
detecting a candidate object in the scene based on the thermal images;
in response to detecting the candidate object, determining whether the candidate object corresponds to an object of interest or a spurious object; and
selectively operating one or more devices in response to determining the candidate object corresponds to the object of interest, wherein the one or more devices comprise a light source and a camera, wherein the selectively operating comprises:
selectively operating the light source to illuminate the scene in response to determining the candidate object corresponds to the object of interest; and
activating the camera to capture images of the scene in response to determining the candidate object corresponds to the object of interest.

US Pat. No. 10,970,555

DATA-DRIVEN EVENT DETECTION FOR COMPRESSED VIDEO

1. A method comprising:obtaining, by a system comprising a processor, a labelled data set comprising historic video data having a plurality of labelled events;
dividing, by the system, the labelled data set into a historic training data set and a historic testing data set;
determining, by the system, using the historic training data set, a plurality of different parameter configurations to be used by a video encoder to encode a video comprising a plurality of video frames, wherein each parameter configuration of the plurality of different parameter configurations comprises a group of pictures (“GOP”) size and a scenecut threshold;
calculating, by the system, an accuracy of event detection and a filtering rate for each parameter configuration of the plurality of different parameter configurations;
calculating, by the system, for each parameter configuration of the plurality of different parameter configurations, a harmonic mean between the accuracy of event detection and the filtering rate; and
selecting, by the system, a best parameter configuration of the plurality of different parameter configurations, wherein the best parameter configuration is the harmonic mean comprising a highest value.

US Pat. No. 10,970,554

METHOD AND SYSTEM FOR AUTOMATICALLY PRODUCING VIDEO HIGHLIGHTS

PIXELLOT LTD., Petach Ti...

1. A method for automatically producing highlight videos from one or more video streams of a playing field, comprising:receiving calibrated raw inputs captured by at least one camera positioned proximate to the playing field, wherein the calibrated raw inputs include audio, calibrated video and actual event time and wherein the audio and the calibrated video are synchronized;
extracting features to transform the calibrated raw inputs into feature vectors by applying low-level processing, wherein said low-level processing comprises utilizing pre-existing knowledge regarding points within the field of view of said at least one camera and therefrom identifying and extracting features, the pre-existing knowledge including: knowledge about areas of the playing field, knowledge about certain players, and knowledge about how various players move around the playing field;
creating segments from the feature vectors and identifying specific events in each one of the segments by applying rough segmentation;
determining whether each one of the events is a highlight by applying analytics algorithms; and,
outputting the highlights for consumption and determining types of packages to be outputted for consumption.

US Pat. No. 10,970,553

SEMANTIC SEGMENTATION OF THREE-DIMENSIONAL DATA

UATC, LLC, San Francisco...

1. A computer-implemented method of semantic segmentation, the method comprising:obtaining, by a computing system comprising one or more computing devices, sensor data comprising three-dimensional data associated with an environment, wherein the three-dimensional data comprises a plurality of points;
determining, by the computing system, data indicative of a two-dimensional voxel representation associated with the environment based at least in part on the three-dimensional data, wherein determining the data indicative of the two-dimensional voxel representation associated with the environment based at least in part on the three-dimensional data comprises
generating a voxel grid based at least in part on the three-dimensional data, the voxel d represented by two horizontal axes and a vertical axis,
generating the two-dimensional voxel representation based at least in part on the voxel grid, wherein the two-dimensional voxel representation is associated with one or more channels;
accessing, by the computing system, a machine-learned semantic segmentation model;
providing as an input, by the computing system, the data indicative of the two-dimensional voxel representation associated with the environment into the machine-learned semantic segmentation model; and
obtaining, by the computing system, an output from the machine-learned semantic segmentation model, wherein the output is indicative of a classification for each point of the plurality of points.

US Pat. No. 10,970,552

SCENE CLASSIFICATION FOR IMAGE PROCESSING

GoPro, Inc., San Mateo, ...

1. A non-transitory computer-readable storage medium, comprising executable instructions that, when executed by a processor, facilitate performance of operations, the operations including operations to:generate, by an image signal processor, a previously processed image, wherein generating the previously processed image includes generating corresponding image processing information;
obtain, from the image signal processor, the image processing information corresponding to the previously processed image;
obtain scene classification information for an input image based on the image processing information, wherein to obtain the scene classification information includes to:
obtain color component ratio information corresponding to a pixel from the previously processed image;
obtain a pixel classification for the pixel by classifying the pixel based on the color component ratio information corresponding to the pixel; and
classify the scene based on the pixel classification;
generate a processed image by processing the input image based on the scene classification information; and
output the processed image.

US Pat. No. 10,970,551

CONTROL APPARATUS AND CONTROL METHOD FOR DETERMINING RELATION OF PERSONS INCLUDED IN AN IMAGE, AND STORAGE MEDIUM STORING A PROGRAM THEREFOR

Canon Kabushiki Kaisha, ...

1. A control apparatus comprising:a controller including a processor and a memory, wherein the controller is configured to perform operations including:
detecting one or more persons included in a moving image,
setting, on the moving image, an area where a plurality of persons is required to work,
determining, as determination result, whether or not a number of detected persons satisfies a notification condition for notifying predetermined information about a person in the set area, and
performing control about notification of the predetermined information in accordance with the determination result,
wherein the notification condition is that a number of persons detected in the set area is one,
wherein, if it is determined that the notification condition is satisfied in the set area, performing control includes performing control to notify, as the predetermined information, information recognizable by a user that there is a person working alone in the set area where the plurality of persons is required to work, and
wherein, if it is determined that the notification condition is not satisfied in the set area, performing control includes performing control not to notify the predetermined information due to detection of a plurality of persons in the set area.

US Pat. No. 10,970,550

SYSTEMS AND METHODS FOR STREAM RECOGNITION

Geenee GmbH, Berlin (DE)...

1. A training system comprising:a hierarchical classifier configured to determine whether image data received from a user device includes a match to one of a plurality of triggers, the match comprising an object wherein:
the hierarchical classifier includes:
a segmentation model configured to generate a segment of the image data including a potential trigger;
a general classifier configured to associate the segment with an identified one of a plurality of general classes; and
a specific classifier corresponding to the identified general class and configured to associate the segment with the one of the plurality of triggers; and
the hierarchical classifier is configured to determine that the image data includes the match when the association between the segment and the one of the plurality of triggers satisfies a match condition;
a secondary classifier configured to determine whether the image data includes the match;
a pose estimation model corresponding to the specific classifier, wherein the system is configured to generate training data using the pose estimation model, in response to a determination by the hierarchical classifier that the image data includes the match, by tracking the object in a video stream including the image data; and
a trainer configured to:
train the specific classifier using the training data, and
train the specific classifier using the segment when the secondary classifier determines that the image data includes the match and the hierarchical classifier determines that the image data does not include the match.

US Pat. No. 10,970,549

VIRTUAL ASSISTANT OF SAFE LOCKER

WELLS FARGO BANK, N.A., ...

1. A method, comprising:providing a locker that is accessible via a portal, wherein the locker facilitates storage of a plurality of items;
capturing an image of the plurality of items;
evaluating at least one of the plurality of items placed inside the locker for digitized metrics wherein the evaluating comprises:
scanning at least one item of the plurality of items; and
determining a value of the at least one item based on the scanning as one of the digitized metrics; and
providing the portal for the locker such that an owner of the locker has exclusive access to the image of the plurality of items and the digitized metrics;
performing augmented reality analysis on the image, wherein the augmented reality analysis includes:
identifying items within the locker using recognition algorithms;
associating the items with identified labels:
annotating the captured image with the identified labels, wherein the annotations are overlaid onto the captured image to create an augmented reality image of the locker
receiving a privacy level selection;
determining items to be redacted from the annotated image based on the privacy level selection; and
redacting one or more items within the annotated image such that the one or more items do not appear in the annotated image.

US Pat. No. 10,970,548

VIRTUAL ASSISTANT OF SAFE LOCKER

WELLS FARGO BANK, N.A., ...

1. A method, comprising:capturing an image of an interior of a locker that includes a plurality of items;
generating an augmented reality overlay on the captured image that annotates each of the plurality of items with identification labels in the image using recognition algorithms;
receiving a privacy level selection out of a set of privacy levels, wherein the privacy level selection determines one or more items of the plurality of items to be redacted from the captured image based on the identification labels;
redacting areas within the captured image via the augmented reality overlay based on the identification labels and the privacy level selection; and
providing access to the captured image and the augmented reality overlay via a portal.

US Pat. No. 10,970,547

INTELLIGENT AGENTS FOR MANAGING DATA ASSOCIATED WITH THREE-DIMENSIONAL OBJECTS

Microsoft Technology Lice...

16. A system for managing data associated with real-world objects and virtual objects displayed in a communication session, comprising:means for receiving sensor data generated by a depth map sensor and an image sensor, the sensor data comprising image data of real-world objects and depth map data defining a three-dimensional model of the real-world objects positioned within a real-world environment;
means for receiving model data defining virtual objects for display within a view of the real-world environment;
means for generating data defining a plurality of agent instances the individual agent instances having an associated database for individual virtual objects and individual real-world objects of interest that are identified by an analysis of the depth map data and the image data;
means for monitoring communication data of a plurality of users of the communication session for identifying keywords, descriptions, or parameters to be stored within individual databases of the individual agent instance associated with the individual virtual objects or the individual real-world objects; and
means for generating, based on the keywords, descriptions, or parameters stored within the individual databases of the individual agent instances associated with the individual virtual objects or the individual real-world objects, a display of one or more recommendations associated with a rendering of the individual virtual objects and a view of at least one real-world object, wherein the one or more recommendations include at least one of a modification to at least one virtual object comprising at least one new parameter, a new virtual object positioned over a real-world object, or an annotation recommending a modification to the at least one real-world object.

US Pat. No. 10,970,546

METHOD AND APPARATUS FOR PROVIDING INFORMATION REGARDING VIRTUAL REALITY IMAGE

Samsung Electronics Co., ...

1. A method of providing information regarding a virtual reality (VR) image, the method comprising:providing the VR image;
obtaining a summation value of polar coordinates of the VR image corresponding to a gaze of a user viewing the VR image;
calculating a change amount over time of the summation value of the polar coordinates;
determining an image corresponding to an area of the VR image in which the gaze of the user is located during a time when the change amount is less than or equal to a first threshold value as a static image of interest;
determining at least one image corresponding to at least one area of the VR image in which the gaze of the user is located during a time when the change amount is between the first threshold value and a second threshold value as a dynamic image of interest; and
providing information regarding at least one at least one of the static image of interest and the dynamic image of interest.

US Pat. No. 10,970,545

GENERATING AND SURFACING AUGMENTED REALITY SIGNALS FOR ASSOCIATED PHYSICAL ITEMS

Amazon Technologies, Inc....

1. A computer-implemented method for displaying augmented reality cues, comprising:receiving, from an imaging device associated with a guide device, video data along a path of an augmented reality experience, the augmented reality experience being presented to a user via a user device;
analyzing the video data to determine visual features associated with the video data;
determining, using a trained computer vision algorithm, that the visual features associated with the video data are consistent with a recognized location along the path of the augmented reality experience;
identifying an augmented reality cue associated with the recognized location, the augmented reality cue further associated with at least one physical item sample provided to the user of the user device prior to initiation of the augmented reality experience, and the at least one physical item sample corresponding to an item available from the recognized location during the augmented reality experience; and
rendering the augmented reality cue for presentation to the user via the user device as part of the augmented reality experience.

US Pat. No. 10,970,544

AUTOMATED WAREHOUSE

Murata Machinery, Ltd., ...

1. An automated warehouse including a plurality of transfer destinations at which articles can be placed and a transporter that moves between the plurality of transfer destinations and transfers the article to the transfer destination, the automated warehouse comprising:a moving-side image capturer provided on the transporter and image-captures the transfer destination and a part or all of an operation of transferring the article to the transfer destination performed by the transporter; and
a fixed-side image capturer provided at a predetermined position in the automated warehouse, and image-captures, from a direction different from that of the moving-side image capturer, the transporter and a part or all of the operation of transferring the article to the transfer destination performed by the transporter,
wherein the transporter includes: a main body that moves between the plurality of transfer destinations; and a transferer provided on the main body and delivers or receives the article to or from the transfer destination,
the moving-side image capturer is provided on the main body while being oriented in a direction in which the transferer delivers or receives the article, and image-captures the interior of each transfer destination,
the fixed-side image capturer is provided while being oriented toward the interior of the automated warehouse from each traveling end side of the transporter, and image-captures the transporter moving between the plurality of transfer destinations,
the fixed-side image capturer is arranged on an upper side of the automated warehouse and a traveling end side of the transporter, and image-captures, from diagonally above, an interior of the automated warehouse including the transporter, and
an image area of the moving-side image capturer and an image area of the fixed-side image capturer overlap.

US Pat. No. 10,970,543

DISTRIBUTED AND SELF-VALIDATING COMPUTER VISION FOR DENSE OBJECT DETECTION IN DIGITAL IMAGES

Accenture Global Solution...

1. A system comprising:a database;
a memory for storing a multilayer landmark recognition and segmentation convolutional neural network; and
system circuitry in communication with the database and the memory, and configured to:
receive aerial images and land maps corresponding to a geographic region from the database;
divide the aerial images into a plurality of partially overlapping aerial blocks and divide the land maps into a plurality of partially overlapping map blocks corresponding to the aerial blocks;
identify a first set of landmarks, the identified first set of landmarks belonging to one of a predetermined set of landmark types in the aerial blocks and store corresponding boundary boxes, locations in the aerial blocks, and landmark types for the first set of landmarks as a first set of training labels;
train an intermediate multilayer convolutional neural network to output a second set of predicted labels containing boundary boxes, locations, and landmark types of landmarks using the aerial blocks and the first set of training labels as input, wherein the intermediate multilayer convolutional neural network is trained to generate a collective error between the first set of training labels and a subset of the second set of predicted labels corresponding to the first set of training labels that are smaller than a first predetermined threshold;
obtain a third set of predicted and validated labels among the second set of predicted labels that are additional to the first set of training labels;
combine the third set of predicted and validated labels and the first set of training labels to form a fourth set of training labels;
train the multilayer landmark recognition and segmentation convolutional neural network to output a fifth set of predicted labels containing boundary boxes, locations, and landmark types of landmarks using the aerial blocks and the fourth set of training labels as input, wherein the multilayer landmark recognition and segmentation convolutional neural network is trained to generate a collective error between the fourth set of training labels and the fifth set of predicted labels that are smaller than a second predetermined threshold; and
input unlabeled aerial blocks that are partially overlapping into the multilayer landmark recognition and segmentation convolutional neural network to obtain predicted landmark labels for the unlabeled aerial blocks.

US Pat. No. 10,970,542

SCALABLE THREE DIMENSIONAL OBJECT SEGMENTATION

HERE Global B.V., Eindho...

1. A method for segmentation of three dimensional objects, the method comprising:identifying point cloud data for a geographic region;
analyzing the point cloud data with a learned model;
assigning probability values outputted from the learned model to points in the point cloud data;
clustering the probability values based on relative locations of the assigned points in the point cloud data;
generating a matrix with at least a portion of the clustered probability values;
factoring the matrix to assign a line for a three dimensional object of the geographic region; and
fitting at least one predetermined shape for the three dimensional object to the point cloud data based at least on the assigned line.

US Pat. No. 10,970,541

GEO-REGISTERING AN AERIAL IMAGE BY AN OBJECT DETECTION MODEL USING MACHINE LEARNING

The Boeing Company, Chic...

1. An apparatus for obtaining and geo-registering an aerial image of an object of interest, the apparatus comprising:a memory configured to store computer-readable program code; and
processing circuitry configured to access the memory, and execute the computer-readable program code to cause the apparatus to at least:
obtain an aerial image by the processing circuitry;
access an object detection model trained using a machine learning algorithm and a training set of aerial images of the object of interest having a defined boundary;
use the object detection model to detect the object of interest in the aerial image, the object detection including a prediction of a boundary of the object of interest depicted in the aerial image based on the defined boundary of the object of interest; and
geo-register the aerial image including the prediction of the boundary of the object of interest with a geographic location of the object of interest.

US Pat. No. 10,970,540

SCHEMES FOR RETRIEVING AND ASSOCIATING CONTENT ITEMS WITH REAL-WORLD OBJECTS USING AUGMENTED REALITY AND OBJECT RECOGNITION

Sony Interactive Entertai...

1. A method comprising:identifying a real-world object in a scene that is being viewed by a camera of a user device;
while the scene is being viewed by the camera of the user device, matching the real-world object with a tagged object based at least in part on image recognition and a sharing setting of the tagged object, the tagged object having been tagged with a content item;
while the scene is being viewed by the camera of the user device, providing a notification to a user of the user device that the content item is associated with the real-world object;
receiving a request from the user for the content item; and
providing the content item to the user.

US Pat. No. 10,970,539

BINDING OF SELFIE FACE IMAGE TO IRIS IMAGES FOR BIOMETRIC IDENTITY ENROLLMENT

Tascent, Inc., Los Gatos...

12. A system comprising:an imager configured to capture a first face image of a subject using light in a visible wavelength range and to capture a second face image of the subject using light in a near infrared wavelength range; and
a processor configured to execute computer-executable instructions for performing a method comprising:
retrieving a selfie face image of a subject, wherein the selfie face image was previously captured using an imaging device that is different from the imager;
matching, by a biometric algorithm, a face portion of the first face image against a face portion of the selfie face image, wherein the face portion for each at least comprises an iris portion;
determining a successful match by the biometric algorithm, the successful match indicating that the subject in the first face image is the same subject as the subject in the selfie face image;
responsive to determining the successful match, segmenting one or more iris portions from the second face image to generate one or more iris images, each iris image corresponding to an iris portion; and
storing in a database an association between (i) the one or more iris images to the (ii) selfie face image, the association indicating that the subject in the one or more iris images is the same subject as the subject in the selfie face image.

US Pat. No. 10,970,538

DYNAMIC BRIGHTNESS AND RESOLUTION CONTROL IN VIRTUAL ENVIRONMENTS

Intel Corporation, Santa...

1. An apparatus, comprising:a graphics processor to dynamically adjust, in response to a measurement of at least one ocular characteristic of an eye of a user that includes a vertical palpebrae position relative to a pupil of the eye of the user, a rendering rate of an image for display between a full rendering and a partial rendering.

US Pat. No. 10,970,537

IMAGE PROCESSING APPARATUS, IMAGE PROCESSING METHOD, AND RECORDING MEDIUM HAVING IMAGE PROCESSING PROGRAM RECORDED THEREIN

OLYMPUS CORPORATION, Tok...

1. An image processing apparatus comprising:a processor that performs a process including
acquiring image information from which a left-eye viewpoint image is capable of being formed and image information from which a right-eye viewpoint image is capable of being formed, so as to generate the left-eye and right-eye viewpoint images,
detecting a plurality of pairs that are each a pair of a partial image of the left-eye viewpoint image and a partial image of the right-eye viewpoint image that are similar to each other, and
performing, for each of the plurality of pairs, image adjustment processing for adjusting three-dimensionality of one of or both of the partial images of the left-eye and right-eye viewpoint images;
a display apparatus that displays the left-eye viewpoint image or the left-eye viewpoint image after the image adjustment processing in a manner such that this image is capable of being observed with a left eye, and displays the right-eye viewpoint image or the right-eye viewpoint image after the image adjustment processing in a manner such that this image is capable of being observed with a right eye, and
an image sensor having a plurality of pixel elements, wherein
for each of the plurality of detected pairs determined on the basis of a positional relationship between the partial images of the left-eye and right-eye viewpoint images that belong to the detected pair, the processor performs the image adjustment processing for one of or both of the partial images of the left-eye and right-eye viewpoint images in such a manner that the positional relationship is changed,
the plurality of pixel elements each include one microlens and a plurality of photoelectric conversions elements, and
the image information from which the left-eye viewpoint image is capable of being formed and the image information from which the right-eye viewpoint image is capable of being formed are acquired from the image sensor.

US Pat. No. 10,970,536

METHOD AND SYSTEM FOR ASSESSING SIMILARITY OF DOCUMENTS

Open Text Corporation, W...

1. A system for assessing similarity of documents, comprising:a processor;
a document repository including an archived document and an archived document metadata vector for archived document metadata, wherein the archived document metadata vector was created by tokenizing the archived document metadata and vectorizing the tokenized metadata to obtain the archived document metadata vector;
a non-transitory computer readable medium, comprising instructions for:
extracting a reference document text from a reference document;
quantifying the reference document by:
tokenizing a portion of the reference document and vectorizing the tokenized portion to obtain a reference document text vector for the portion of the reference document, and
tokenizing reference document metadata and vectorizing the tokenized metadata to obtain a reference document metadata vector for the reference document metadata;
associating at least a portion of the quantified reference document and the archived document based on the reference document metadata and the archived document metadata vector; and
linking the at least the portion of the quantified reference document and the archived document based on the association between the at least the portion of the quantified reference document and the archived document.

US Pat. No. 10,970,535

SYSTEM AND METHOD FOR EXTRACTING TABULAR DATA FROM ELECTRONIC DOCUMENT

Innoplexus AG, Eschborn ...

1. A system for extracting tabular data from an electronic document, the system having a data processing arrangement that is to:(i) receive the electronic document;
(ii) determine a location of the tabular data within the electronic document;
(iii) extract an image of the tabular data from the electronic document;
(iv) convert the received image of the tabular data into a greyscale image of the tabular data;
(v) extract a grid structure of the tabular data from the greyscale image;
(vi) remove the grid structure of the tabular data from the greyscale image, by superimposing the extracted grid structure on to the greyscale image;
(vii) determine a position for placement of horizontal and vertical lines in the greyscale image without the grid structure;
(viii) generate horizontal and vertical lines on the greyscale image without the grid structure, to obtain information of grid structure of the tabular data;
(ix) perform optical character recognition of the text associated with the tabular data from the received image, to obtain information of text associated with the tabular data; and
(x) extract the tabular data by combining the information of the grid structure of the tabular data with the text associated with the tabular data, to generate the tabular data.

US Pat. No. 10,970,534

DOCUMENT PROCESSING SYSTEM CAPTURE FLOW COMPILER

OPEN TEXT CORPORATION, W...

1. A system comprising:a communication interface;
a processor coupled to the communications interface; and
a non-transitory computer readable medium coupled to the processor and storing a set of computer executable instructions executable by the processor to:
receive an input capture flow comprising a first sequence of steps representing configurable code components of a document capture system, wherein a step in the first sequence of steps holds configuration information for a corresponding configurable code component;
compile the input capture flow into a capture process that defines a processing order for the configurable code components and setup attribute values for the configurable code components, wherein compiling the input capture flow into the capture process comprises determining an optimized second sequence of steps that comprises steps from the first sequence of steps and compiling the second sequence of steps into the capture process, and wherein determining the second sequence of steps comprises:
reordering, for parallel execution, a group of independent steps indicated in the input capture flow for sequential execution; and
based on a determination that the input capture flow contains a duplicative step, eliminating the duplicative step, wherein compiling the input capture flow into the capture process comprises building an in-memory model of the input capture flow and manipulating the in-memory model to reorder the group of independent steps and eliminate the duplicative step; and
deploy the capture process to the document capture system.

US Pat. No. 10,970,533

METHODS AND SYSTEMS FOR FINDING ELEMENTS IN OPTICAL CHARACTER RECOGNITION DOCUMENTS

INTERNATIONAL BUSINESS MA...

1. A method for finding elements in optical character recognition (OCR) documents comprising:receiving an indication of a selected portion of document;
determining salient pixels in the selected portion of the document;
identifying properties of the salient pixels in the selected portion of the document;
comparing the properties of the salient pixels in the selected portion of the document to properties of pixels in each of a plurality of portions of an OCR-converted version of the document; and
utilizing a cognitive analysis to select at least some of the plurality of portions of the OCR-converted version of the document as suspected matches to the selected portion of the document.

US Pat. No. 10,970,532

SYSTEMS AND METHODS FOR PREVENTING SHARING OF SENSITIVE CONTENT IN IMAGE DATA ON A CLOSED COMPUTING PLATFORM

CA, Inc., San Jose, CA (...

1. A computer-implemented method for preventing sharing of sensitive content in image data on a closed computing platform, at least a portion of the method being performed by one or more computing devices comprising at least one processor, the method comprising:detecting, by the one or more computing devices, initiation of a network connection for sending network traffic data to a data storage service on the closed computing platform;
monitoring, by the one or more computing devices, the sending of the network traffic data to identify a target traffic indicator associated with the image data;
interrupting, by the one or more computing devices, the sending of the network traffic data upon identifying the target traffic indicator;
analyzing, by the one or more computing devices, the image data to identify the sensitive content; and
performing, by the one or more computing devices, a security action that protects against the sensitive content being shared to the data storage service on the closed computing platform.

US Pat. No. 10,970,531

DIGITIZATION OF INDUSTRIAL INSPECTION SHEETS BY INFERRING VISUAL RELATIONS

Tata Consultancy Services...

1. A processor implemented method (200) comprising the steps of:receiving, by one or more hardware processors, an input comprising a plurality of images, the plurality of images being characterized by identical one or more templates, wherein each template in the one or more templates comprises a static part being a line diagram representation of a machine having one or more predefined zones and a dynamic part comprising handwritten content associated with the one or more predefined zones of a corresponding machine (202);
eliminating, by the one or more hardware processors, the static part from the one or more templates from each of the plurality of images to obtain a plurality of de-templated images comprising the dynamic part, wherein the handwritten content comprises one or more independent text patches and one or more dialogue clouds surrounding the independent text patches, associated with one or more connectors (204);
processing, by the one or more hardware processors, the plurality of de-templated images to eliminate the one or more dialogue clouds thereof to obtain a plurality of de-clouded images using an encoder-decoder based SegNet architecture for segmenting the one or more dialogue clouds, the de-clouded images comprising a plurality of text patches (206), wherein the SegNet architecture is pre-trained on a dataset of a plurality of dialogue cloud images;
detecting, by the one or more hardware processors, the one or more connectors in the plurality of de-clouded images by using a Convolutional Neural Network (CNN) classifier and a Hough Line Transform to detect at least one of one or more arrows and one or more lines respectively that constitute the one or more connectors (208);
detecting, by the one or more hardware processors, coordinates of each text patch in each of the plurality of images of the input using a Connectionist Text Proposal Network (CTPN), the coordinates forming bounding boxes around each text patch (210);
mapping each of the one or more connectors, by the one or more hardware processors, to a corresponding text patch based on the coordinates associated thereof and by using a clustering method (212);
identifying, by the one or more hardware processors, text associated with each text patch, in American Standard Code for Information Interchange (ASCII) format, using the de-clouded image, a Capsule Network (CapsNet) and a Spatial Transformer Network (STN) (214), wherein identifying the text associated with each text patch comprises:
segmenting each text patch using a Connected Component Analyses (CCA) to generate segments having one or more characters therein;
ranking the segmented characters in the generated segments that are unordered to obtain characters arranged in a human readable form;
using the CapsNet to recognize the generated segments having more than one characters therein; and
using the STN to recognize the generated segments having one character therein, upon the CapsNet predicting ‘None’ as an output, when only one character is present in the generated segments; and
performing, by the one or more hardware processors, one-to-one mapping of the identified text associated with each text patch to the one or more predefined zones of the corresponding machine, thereby providing a visual relation therebetween, by using the mapped one or more connectors and the coordinates of the corresponding text patch (216).

US Pat. No. 10,970,530

GRAMMAR-BASED AUTOMATED GENERATION OF ANNOTATED SYNTHETIC FORM TRAINING DATA FOR MACHINE LEARNING

Amazon Technologies, Inc....

14. A system comprising:a storage service implemented by a first one or more electronic devices; and
a training data generation engine implemented by a second one or more electronic devices, the training data generation engine including instructions that upon execution cause the training data generation engine to:
generate a plurality of documents, wherein for each document of the plurality of documents, the training data generation engine is to:
determine, based on a defined grammar, a plurality of sections for the document,
select, for each of the plurality of sections, one or more key-value units, and
select, for each of the selected one or more key-value units, one or more style attributes and place the one or more key-value units within the section; and
store the plurality of documents to a storage location provided by the storage service.

US Pat. No. 10,970,529

VEHICLE IMAGING SYSTEM, VEHICLE IMAGING METHOD AND DEVICE, PROGRAM, AND RECORDING MEDIUM

NEC CORPORATION, Tokyo (...

1. A vehicle window detection system comprising:a displacement sensor configured to emit a laser beam into a roadway space, and receive a reflected beam reflected by the vehicle present in the roadway space; and
one or more processors that are configured to:
utilize the displacement sensor to measure a first distance from the displacement sensor to a first reflection point where the laser beam is reflected by a body of the vehicle without being transmitted through a window of the vehicle and measure a second distance from the displacement sensor to a second reflection point where the laser beam is transmitted through the window of the vehicle and is reflected by an interior of the vehicle;
obtain a first horizontal distance from a position of the displacement sensor to a position of the first reflection point in the horizontal direction based on the first distance, and a second horizontal distance from the position of the displacement sensor to a position of the second reflection point in the horizontal direction based on the second distance; and
detect a window of the vehicle based on a comparison between the first horizontal distance and the second horizontal distance,
wherein the one or more processors is further configured to, prior to detecting the window of the vehicle, detect the vehicle is present in the roadway space based on a change in a vertical distance from the position of the displacement sensor in a vertical direction to a position of the reflection point in the vertical direction between the first reflection point and the second reflection point, and detect the window of the vehicle after the vehicle has been detected.

US Pat. No. 10,970,528

METHOD FOR HUMAN MOTION ANALYSIS, APPARATUS FOR HUMAN MOTION ANALYSIS, DEVICE AND STORAGE MEDIUM

BAIDU ONLINE NETWORK TECH...

1. A method for human motion analysis, comprising:acquiring image information captured by a plurality of photographing devices, wherein at least one of the plurality of photographing devices is disposed above a shelf;
performing human tracking according to the image information captured by the plurality of photographing devices, and determining position information in space of at least one human body and identification information of the at least one human body;
acquiring, according to the position information in space of a target human body of the at least one human body, a target image captured by the photographing device above a shelf corresponding to the position information;
acquiring a key point of the target human body in the target image, wherein the key point of the target human body comprises a key point of a hand of the target human body; and
recognizing an action of the target human body according to the key point of the hand of the target human body and detection data of a non-visual sensor corresponding to the position information;
wherein the non-visual sensor comprises a light curtain sensor which is disposed on a beam of the shelf facing a customer and used to detect an occurrence time of the action and an occurrence position of the action, and an infrared sensor that is disposed on a layer plate of the shelf; and
the recognizing an action of the target human body according to the key point of the hand of the target human body and detection data of a non-visual sensor corresponding to the position information comprises:
acquiring, according to the occurrence position of the action detected by the light curtain sensor, an intensity change of infrared radiation of a commodity detected by the infrared sensor corresponding to the occurrence position of the action at the occurrence time of the action detected by the light curtain sensor; and
recognizing, according to the intensity change of infrared radiation of the commodity detected by the infrared sensor corresponding to the occurrence position of the action and the key point of the hand of the target human body, the action of the target human body.