Amazon Technologies, Inc.

Netbase Solutions, Inc.

Cisco Technology, Inc.

Flextronics Ap, LLC

Zoom Video Communications, Inc.

Logitech Europe S.a.

Apple Inc.

Logitech, Inc.

Wei Li

Michael Jacob Osofsky

Lokesh Pooranmal Bajaj

Wei Xia

Jens Erik Tellefsen

Ranjeet Singh Bhatia

Yuanjun Xiong

Shuo Yang

Fuling Liu

Bill Reckwerdt

Meng Wang

Lin Han

Samuel Waising Tam

Wei-Lien Hsu

Eric Chen

Tak Shing Pang

Yifan Xing

Siqi Deng

Leo Benedict Baldwin

Hao Pan

Amit Kumar Agrawal

Haitao Guo

Jacobi Colton Grillo

Daniel Bibireata

Tony David

Ambrish Tyagi

Chang Yuan

Remy Zimmermann

Geoffrey Heller

Timothy Theodore List

Stefano Soatto

Krishnakanth Rapaka

Xiaohua Yang

Dongqing Cao

Bryed L Billerbeck

Haiyan He

Sergio Maggi

Eddie Alex Azuma

Richard Howard Suplee, Iii

Munehiro Nakazato

David Wilson Poole

Honghong Peng

Mark Hsu

Linfeng Guo

George Sanchez

Xiang Fu

Xu Zhao

Mitchell Norcross

Xilin Li

Xinmin Yan

Yantao Shen

Nicholas Jialei Su

Georges Auberger

Jean-Michel Junien Labrousse

Techniques are presented for producing demographics, in an automated fashion, from a search result of computer-accessible content. While the demographics can be determined for a research product that has been produced by any technique, they are particularly useful when applied to an automated frame-based search approach. Frame-based search engines are presented for technology profiling, healthcare-related search and brand research. Determination of a demographic proceeds at two levels: member and population. At the member level, a demographic characteristic can be determined applicable with either total or partial certainty. Each value assigned by a demographic, to a population member, has a confidence level associated with it and the assignments can be represented by a Confidence Distribution. Summarization of a demographic, at the population level, depends upon whether the certainty assignments, at the member level, are total or partial. Declarant Demographics are presented. Approaches, to determining Declarant Demographics, are presented.

Method and apparatus for determining search result demographics

Methods and systems for transmitting content during a networked conference. In an embodiment, a method is provided for secure access to online events. The method includes receiving a request from a user to access an online event, obtaining a user email address, an event identifier and an event passcode, and hashing the user email address, the event identifier and the event passcode to generate a ticket hash value. The method also includes forming a secure ticket that comprises the ticket hash value, and transmitting the secure ticket to the user. The method also includes receiving the secure ticket from the user when the user want to access the event, hashing the user email address, the event identifier, and the event passcode to generate a confirming hash value, and comparing the two hash values and granting event access to the user if the two hash values match.

Methods and apparatus for secure access to online events

Voice-controlled devices that include one or more speakers for outputting audio. In some instances, the device includes at least one speaker within a cylindrical housing, with the speaker aimed or pointed away from a microphone coupled to the housing. For instance, if the microphone resides at or near the top of the cylindrical housing, then the speaker may point downwards along the longitudinal axis of the housing and away from the microphone. By pointing the speaker away from the microphone, the microphone will receive less sound from the speaker than if the speaker were pointed toward the microphone). Because the voice-controlled device may perform speech recognition on audio signals generated by the microphone, less sound from the speaker represented in the audio signal may result in more accurate speech recognition, and/or a lesser need to perform acoustic echo cancelation (AEC) on the generated audio signals.

Audio assemblies for electronic devices

A frame represents a concept with a set of roles and a set of linguistic rules. If a linguistic rule is satisfied, by a unit of natural language discourse (UNLD), the frame is invoked and a frame instance produced. A frame instance specifies how the UNLD, with particular values drawn from the UNLD, fulfills the roles of the frame. A frame-based search, of target content, is accomplished in response to a frame-based user query. The search result is comprised of records, where each record is a result of a match, of the frame-based query, at a location in the target content. If the frame-based query is implicit, a match requires only that the location of the target content invokes the appropriate frame. If the frame-based query is role-specific, in addition to invoking the appropriate frame, a query search term needs to be found in the value for a role of the frame instance produced.

Method and apparatus for frame-based search and analysis

One example method for conducting a conference between conference participants includes obtaining a scene layout for the conference, the scene layout comprising a plurality of video areas that are each assigned to a respective display video stream from one of the conference participants, and the scene layout forming a common visual presentation for the conference; receiving video streams from one or more of the conference participants; and displaying the scene layout, wherein the video streams from the conference participants are displayed in their respective assigned video areas.

Scene layouts in video conferences

An image may be evaluated by a computer vision system to determine whether it is fit for analysis. The computer vision system may generate an embedding of the image. An embedding quality score (EQS) of the image may be determined based on the image's embedding and a reference embedding associated with a cluster of reference noisy images. The quality of the image may be evaluated based on the EQS of the image to determine whether the quality meets filter criteria. The image may be further processed when the quality is sufficient, or otherwise the image may be removed.

Methods and systems for performing noise-resistant computer vision techniques

Techniques for data synthesis for training datasets for machine learning applications are described. A first image of at least an object from a first viewpoint is obtained. The first image having associated first image metadata including a first location of a feature of the object in the first image. A model is generated from the first image, the model including a three-dimensional representation of the object. A second image is generated from the model, the second image including the object from a second viewpoint that is different from the first viewpoint. Second image metadata is generated, the second image metadata including a second location of the feature of the object in the second image, the second location corresponding to the first location adjusted for the difference between the second viewpoint and the first viewpoint.

Data synthesis using three-dimensional modeling

Systems, methods, and devices are disclosed for detecting an active speaker in a two-way conference. Real time audio in one or more sub band domains are analyzed according to an echo canceller model. Based on the analyzed real time audio, one or more audio metrics are determined from output from an acoustic echo cancellation linear filter. The one or more audio metrics are weighted based on a priority, and a speaker status is determined based on the weighted one or more audio metrics being analyzed according to an active speaker detection model. For an active speaker status, one or more residual echo or noise is removed from the real time audio based on the one or more audio metrics.

Acoustic echo cancellation based sub band domain active speaker detection for audio and video conferencing applications

Techniques for building a backward compatible and backfill-free image search system are described. According to some embodiments, a backwards compatible training system trains a new embedding model to be backward compatible with the face embeddings (e.g., floating-point vectors) generated by a previous embedding model. In one embodiment, backwards compatible training uses a classifier of the previous embedding model as a form of constraint in the training of the new embedding model.

Backward compatible and backfill-free image search system

Methods and system for providing scene layouts during a network conference. In an embodiment, a method for providing a scene layout for use by conference participants during a network conference is disclosed. The method includes identifying a conference host and conference attendees from the conference participants, identifying a scene layout selected by the conference host, and distributing the scene layout to the conference participants. The scene layout includes a plurality of video areas. The method also includes assigning selected video areas of the scene layout to selected conference participants, respectively, receiving video streams from the selected conference participants, and displaying the scene layout so that the video streams from the selected conference participants are displayed in their associated video areas, respectively.

Methods and system for providing scene layouts during a network conference

Method and apparatus for frame based search

Systems, methods, and devices are disclosed for detecting an active speaker in a two-way conference. Real time audio in one or more sub band domains are analyzed according to an echo cancellor model. Based on the analyzed real time audio, one or more audio metrics are determined from output from an acoustic echo cancellation linear filter. The one or more audio metrics are weighted based on a priority, and a speaker status is determined based on the weighted one or more audio metrics being analyzed according to an active speaker detection model. For an active speaker status, one or more residual echo or noise is removed from the real time audio based on the one or more audio metrics.

Techniques are disclosed for coding image data adaptively at different levels of downscaling. Such techniques may involve partitioning input data into pixel blocks for coding and performing content analysis on the pixel blocks. The pixel blocks may be input to block coders that operate at different pixel block sizes, which may code the pixel blocks input to them at their respective sizes. Except when a block coder operates at the partitioning size, block coders that operate at different pixel block sizes may perform downscaling of the pixel blocks to match their size with the block coders' respective coding size. A block decoder may invert the coding operations performed by the block coders, decoding coded image data at respective pixel block sizes, then upscaling decoded image data obtained therefrom to a common pixel block size. Image reconstruction may synthesize a resultant image from the decode pixel block data output by the decoders. Quality enhancements may be added by neural network processing systems that have been trained to other segmented images.

Deep quality enhancement of adaptive downscaled coding for image compression

An echo cancellation method includes receiving an echo reference signal, receiving a microphone signal, decomposing, with a first filter bank, the echo reference signal into a series of subband echo reference signals, decomposing, with a second filter bank, the microphone signal into a series of subband microphone signals, estimating a group delay between the echo reference signal and the microphone signal using the series of subband echo reference signals and the series of subband microphone signals, estimating, using adaptive filters, acoustic echoes in the echo reference signal based at least in part on the group delay, subtracting the acoustic echoes from the series of subband microphone signals to obtain a series of acoustic echo removed subband signals, combining the series of acoustic echo removed subband signals into a single time domain echo removed signal, and sending the single time domain echo removed signal to a host operating system.

Acoustic echo cancellation with delay uncertainty and delay change

Growing community of inventors

Cupertino, CA, United States of America

Wei Li