Sony Corporation

Sony Electronics Inc

Sony Group Corporation

Sas Institute Inc.

Saturn Licensing LLC

Robert Winston Blanchard

Peter Rae Shintani

Brant L Candelore

Mahyar Nejat

Mark Eyer

Robert Hardacker

Adam Michael Goldberg

Christopher Jensen Read

Leo M Pedlow, Jr

Rolf Toft

Toshiro Ozawa

George Edward Williams

David Boyden

Ruiwen Zhang

Xudong Yu

Neela Niranjani Vengateshwaran

Steven Richman

Behram Mario Dacosta

Jorge Manuel Gomes Da Silva

Varunraj Valsaraj

Yang Yu

Mohammadreza Nazari

Aravind Asam

Prathaban Mookiah

Behram daCosta

Takashi Hironaka

Amirhassan Fallah Dizche

Weichen Wang

Jonathan Lee Walker

Koichi Oshima

William Robert Nadolski

Afshin Oroojlooyjadid

Davood Hajinezhad

Hardi Desai

Hamoon Azizsoltani

Ye Liu

Hoda Sayyadinejad

Vrushali Sawant

Jean Baronas

Mark Kenneth Eyer

Mahyar Mike Nejat

A system and method include generating synthetic data by generating a first set of hyperparameters for a first trained machine learning model and a second set of hyperparameters for a second trained machine learning model, generating a plurality of synthetic data vectors using the first and second trained machine learning models, computing an error function for the first and second set of hyperparameters using a third machine learning model, computing an objective function value, responsive to determining that the objective function value is not an optimal value, updating the first set of hyperparameters and the second set of hyperparameters or responsive to determining that the objective function value is an optimal value outputting the plurality of synthetic data vectors as a set of synthetic data.

Techniques for generating synthetic data

A system, method, and computer-program product includes detecting, via a localization machine learning model, a target object within target image data of a scene, classifying, via an object classification machine learning model, the target object to a probable object class of a plurality of distinct object classes, routing, via the one or more processors, the target image data of the scene to a target object-condition machine learning classification model of a plurality of distinct object-condition machine learning classification models based on a mapping between the plurality of distinct object classes and the plurality of distinct object-condition machine learning classification models, classifying, via the target object-condition machine learning classification model, the target object to a probable object-condition class of a plurality of distinct object-condition classes, and displaying, via a graphical user interface, a representation of the target object in association with the probable object-condition class.

Systems and methods for configuring and using a multi-stage object classification and condition pipeline

An electronic device and method are provided for display of signing video through an adjustable user interface (UI) element. The electronic device receives a first media stream that includes a video. The electronic device determines a location of a signer in the video. The electronic device extracts a video portion that corresponds to the determined location in the video. The electronic device controls a playback of the video on the display device. The electronic device controls the display device based on the playback to render the UI element on the display device and display the extracted video portion inside the UI element.

Display of signing video through an adjustable user interface (UI) element

A system, method, and computer-program product includes detecting, via a localization machine learning model, a target object within a scene based on downsampled image data of the scene, identifying a likely position of the target object within original image data of the scene, extracting, from the original image data of the scene, a target sub-image containing the target object, classifying, via an object classification machine learning model, the target object to a probable object class of a plurality of distinct object classes, routing the target image resolution of the target sub-image to a target object-condition machine learning classification model of a plurality of distinct object-condition machine learning classification models, classifying, via the target object-condition machine learning classification model, the target object to a probable object-condition class, and displaying, via a graphical user interface, a representation of the target object in association with the probable object-condition class.

Provided is an electronic device that captures a first image of a user. The electronic device determines a position of eye gaze of the user on a first UI element of a plurality of UI elements displayed on a first display screen of the electronic device, based on the first image. The electronic device selects the first UI element based on a determination of the position of the eye gaze on the first UI element for a time interval. The electronic device transmits a first control signal to a media display device, based on the selection of the first UI element. The electronic device receives a feedback image from the media display device. The feedback image indicates control of the media display device based on the first control signal. The electronic device controls the first display screen to display the feedback image with the plurality of UI elements.

Media display device control based on eye gaze

A media rendering device and method for reproduction control of scene description is provided. The media rendering device retrieves media content that includes a set of filmed scenes and text information. The text information includes video description information and timing information. The video description information describes a filmed scene in the set of filmed scenes. The media rendering device further extracts the timing information to reproduce the video description information from the text information of the filmed scene. The media rendering device further controls the reproduction of the video description information in either a textual representation or in a textual and audio representation at a first-time interval indicated by the extracted timing information of the filmed scene.

Reproduction control of scene description

A computing device trains a machine state predictive model. A generative adversarial network with an autoencoder is trained using a first plurality of observation vectors. Each observation vector of the first plurality of observation vectors includes state variable values for state variables and an action variable value for an action variable. The state variables define a machine state, wherein the action variable defines a next action taken in response to the machine state. The first plurality of observation vectors successively defines sequential machine states to manufacture a product. A second plurality of observation vectors is generated using the trained generative adversarial network with the autoencoder. A machine state machine learning model is trained to predict a subsequent machine state using the first plurality of observation vectors and the generated second plurality of observation vectors. A description of the machine state machine learning model is output.

Automated control of a manufacturing process

A display device such as a TV has an image capture device to image gestures of a hearing-impaired user. The display device operates in a navigation mode, in which non-sign gestures can be made by the user to navigate a user interface, and a sign mode, in which sign language gestures may be made by the user to enter search terms to search for content.

Using signing for input to search fields

An electronic device includes circuitry, firmware, and software that determines identification information associated with a first performer-of-interest at a live event and retrieves a first set of audio tracks from a plurality of audio tracks based on the determined identification information. The circuitry receives a first audio segment associated with the first performer-of-interest from an audio capturing device. The circuitry compares a first audio characteristic of the first audio segment with a second audio characteristic of a first audio portion of each of the first set of audio tracks. The circuitry determines a first audio track based on the comparison between the first audio characteristic and the second audio characteristic. The circuitry identifies a start position of the first audio track based on the first audio segment associated with the first audio track. The circuitry controls a display of the first lyrics information of the first audio track.

Audio track determination based on identification of performer-of-interest at live event

Implementations generally relate to improving accessibility for digital devices. In some implementations, a method includes receiving a first selection of at least one button of a plurality of buttons of a user interface. The method further includes determining a type of the first selection based on one or more button selection characteristics. The method further includes determining a first command based at least in part on the type of the first selection. The method further includes providing an audible indication corresponding to the first command, wherein the audible indication indicates button information.

Accessibility for digital devices

A display system includes a display device and a server. The server receives a plurality of audio segments from a plurality of audio-capture devices. The server receives a user-input that corresponds to a selection of a first user interface (UI) element that represents a first performer-of-interest or a first audio-capture device attached to the first performer-of-interest. The server detects a second performer-of-interest associated with a second audio-capture device within a threshold range of the first audio-capture device. The server extracts a first audio segment of a first vocal deliverance of the first performer-of-interest and a second audio segment of a second vocal deliverance of the second performer-of-interest. The server deduces new caption information from a first verbatim text that is common between the first audio segment and the second audio segment and controls display of the new caption information on the display device.

Display control of different verbatim text of vocal deliverance of performer-of-interest in a live event

An information processing device includes circuitry that receives a user-input for selection of one of a visual accessibility feature and an aural accessibility feature and further receives a first audio segment from an audio capturing device at a live event. The first audio segment includes a first audio portion of the audio content and a first audio closed caption (CC) information. The circuitry controls display of first text information for the first audio portion and second text information for the first audio CC information, based on received user-input for the selection of the visual accessibility feature. The circuitry generates a second audio segment from the first audio segment based on a first audio characteristic of the first audio segment. The circuitry controls a playback of the generated second audio segment, based on the received user-input for the selection of the aural accessibility feature.

Presentation of audio and visual content at live events based on user accessibility

An electronic apparatus and method for synchronization of wireless audio to video is provided. The electronic apparatus determines a first wireless audio processing delay associated with the electronic apparatus and receives media content comprising video content and audio content associated with the video content. The electronic apparatus transmits the audio content to the wireless audio device and controls playback of the video content on the display device based on the determined first wireless audio processing delay such that the playback of the video content is time-synchronized with playback of the audio content on the wireless audio device.

Synchronization of wireless-audio to video

Implementations generally relate to managing audio and video content blocking. In some implementations, a method includes receiving a media item at a media device. The method further includes accessing content information associated with the media item. The method further includes determining a plurality of content categories based on the content information. The method further includes analyzing the content information based on the plurality of content categories. The method further includes comparing analysis results from the analyzing of the content information to a plurality of objectionability thresholds. The method further includes blocking the media item from being played on the media device if the analysis results exceed one or more objectionability thresholds of the plurality of objectionability thresholds.

Managing audio and video content blocking

An application on a mobile phone or tablet computer or eyeglasses listens to the speech of a profoundly deaf person and interprets it in real time. The voice recognition can be trained to the voice of the profoundly deaf person. A display such as a small tablet display which can be hung around the speaker's neck can present text output by the voice recognition module to allow someone to read the text. The application can then detect speech from the listener and display text converted from the speech to the deaf person through glasses or to a cell phone that is being held by hand or to the tablet, or to a different tablet that the previous listener has hung around his or her neck. In case of the same tablet the text from the listener's speech can be shown upside down so that the deaf person can look down and read the text. Thus, the top of the tablet presents text from speech of the deaf person in the up position, and the bottom portion of the tablet presents text from speech of the listener upside down, to be read by the deaf person.

Duplicate speech to text display for the deaf

Growing community of inventors

San Diego, CA, United States of America

Robert Winston Blanchard