Tencent America LLC

Applied Materials, Inc.

Future Wei Technologies, Inc.

Intuit, Inc.

Intel Corporation

Schering Corporation

Extreme Networks, Inc.

Microsoft Technology Licensing, LLC

Finisar Corporation

Tesla, Inc.

Huawei Technologies Co., Limited

Amazon Technologies, Inc.

Broadcom Corporation

Bayerische Motoren Werke Aktiengesellschaft

Intersil Americas Inc.

Wei W Wang

Shan Liu

Wei Jiang

Xiaozhong Xu

Alexander S Ran

Christopher Z Lesner

Marko Rukonic

Ding Ding

Anantha K Subramani

Haoping Yu

Byeongdoo Choi

Xinliang Lu

Karl M Brown

Vineet Haresh Mehta

Chien-Teh Kao

Thomas J Schall

Ping Peter Xie

Hanbing Wu

Mei Yin Chang

Jianming Fu

Erik De La Iglesia

Nadim I Maluf

Albert Zlotnik

Yu Wen Chang

Semyon Sherstinsky

Joel M Huston

Ashish Goel

See-Eng Phan

Stephan Wenger

Zheng Xu

Hernan Adolfo Castro

John C Forster

Praburam Gopalraja

Cheng-Hsiung Matt Tsai

Handong Li

Kurt C Gish

Brandon Barnett

Gordon D Holt

Allen K Lau

Qiang Wu

Aaron Muir Hunter

Sheng Lin

Keith A Miller

Martin Lee Riker

Xinyu Fu

Xiao Yang

John Pipitone

Daniel J Dairaghi

Ahmed Hassan Awadallah

Narayan Sundararajan

Joseph A Hedrick

Etienne Iliffe-Moon

Srinivas Guggilla

Elmoustapha Ould-Ahmed-Vall

Vijay D Parkhe

Seshadri Ganguli

Jing Ye

Eriko Nurvitadhi

Rongjun Wang

Kirankumar Neelasandra Savandaiah

Ratinder Paul Singh Ahuja

Kartik Bhupendra Shah

Mark E McDonald

James S Papanu

Yang Li

Yu Lei

Roderick Craig Mosely

Amit Bleiweiss

Deborah T Marr

Ananthkrishna Jupudi

Alan A Ritchie

Ralf Hofmann

Bradley Owen Stimson

Shaunak Chatterjee

Xiaoxiong Yuan

Jiunn Benjamin Heng

Joung Joo Lee

Vishwas Kumar Pandey

Wei Guang Lu

Prashanth Kothnur

Mark Naoshi Kawaguchi

Lyndon Kennedy

Susan Elizabeth Carrie

Jeffrey J Cook

Rahoul Puri

Srinivas Sridharan

Chi Hong Ching

Asit K Mishra

Rick Lowe

Dheevatsa Mudigere

Thanh X Nguyen

Michael Sterling Jackson

John Joseph Mazzocco

Xiaodong Wang

Donny Young

Lei Pei

Meng Chen

Kah Kuen Fu

Bingxi Sun Wood

Kevin Dapeng Zhang

A method and system assist users of an electronic bookkeeping system to sort financial transactions into accounts representing bookkeeping categories. The method and system generate vectors representing the accounts and the merchants known to the bookkeeping system. The method and system generate compressed versions of the vectors by compressing the vectors. The method and system assist users to sort the financial transactions by analyzing the compressed vectors.

System and method for dimensionality reduction of vendor co-occurrence observations for improved transaction categorization

Certain aspects of the present disclosure provide techniques for determining similarities between businesses. One example method generally includes receiving a similarity query and receiving transaction data associated with a plurality of businesses for comparing the plurality of businesses. The method further includes generating a set of vectors representing the plurality of businesses based on the transaction data and generating a set of compact vectors based on the vectors by applying a dimensionality reduction technique. The method further includes generating based on the set of compact vectors, a similarity index and determining a response to the similarity query using the similarity index.

Responding to similarity queries using vector dimensionality reduction

An apparatus to facilitate acceleration of machine learning operations is disclosed. The apparatus comprises at least one processor to perform operations to implement a neural network and accelerator logic to perform communicatively coupled to the processor to perform compute operations for the neural network.

Machine learning accelerator mechanism

Aspects of the disclosure provide a method, an apparatus, and a non-transitory computer-readable storage medium for video decoding. The apparatus can include processing circuitry. The processing circuitry is configured to decode first neural network update information in a coded bitstream for a first neural network in the video decoder. The first neural network is configured with first pretrained parameters. The first neural network update information corresponds to a first block in an image to be reconstructed and indicates a first replacement parameter corresponding to a first pretrained parameter in the first pretrained parameters. The processing circuitry is configured to update the first neural network in the video decoder based on the first replacement parameter. The processing circuitry can decode the first block based on the updated first neural network for the first block.

Block-wise content-adaptive online training in neural image compression

A method of task-adaptive pre-processing (TAPP) for neural image compression is performed by at least one processor and includes generating a substitutional image, based on an input image, using a TAPP neural network, and encoding the generated substitutional image to generate a compressed representation, using a first neural network. The TAPP neural network is trained by generating a substitutional training image, based on an input training image, using the TAPP neural network, encoding the generated substitutional training image to generate a compressed training representation, using the first neural network, decoding the generated compressed training representation to reconstruct an output training image, using a second neural network, generating gradients of a rate-distortion (R-D) loss that is generated based on the input training image, the reconstructed output training image and the generated compressed training representation, and updating the generated substitutional training image, based on the generated gradients of the R-D loss.

Method and apparatus for task-adaptive pre-processing for neural image compression

A method of three-dimensional (3D)-Tree coding for neural network model compression, is performed by at least one processor, and includes reshaping a four-dimensional (4D) parameter tensor of a neural network into a 3D parameter tensor of the neural network, the 3D parameter tensor comprising a convolution kernel size, an input feature size, and an output feature size, partitioning the 3D parameter tensor along a plane that is formed by the input feature size and the output feature size into 3D coding tree units (CTU3Ds), partitioning each of the CTU3Ds into a plurality of 3D coding units (CU3Ds) recursively until a predetermined depth, using a quad-tree, and constructing a 3D tree for each of the plurality of CU3Ds, wherein the 3D tree for each of the plurality of CU3Ds is a 3D-Unitree.

Method and apparatus for three-dimensional (3D)-tree coding for neural network model compression

An autonomous mobile device (AMD) follows a user who may move among other people, change orientation, or move in and out view of a camera on the AMD. A predicted location of the user may be determined based on a previous location and previous movement. A proximity value is determined by the distance between a location in physical space of a user and the predicted location. An image from a camera is processed to determine a user depicted in the image and generate a feature vector (FV) of that depiction. A gallery of FVs of the user as viewed from different angles, poses, and so forth is stored. A similarity value is determined between a FV of an unidentified user in an image and the FVs in the gallery. The user may be identified using the proximity value and the similarity value.

System for identifying a user with an autonomous mobile device

Aspects of the disclosure provide a method and an apparatus for video encoding. The apparatus includes processing circuitry configured to perform an iterative update of sample values of a plurality of samples in an initial input image. The iterative update includes generating a coded representation of a final input image based on the final input image by an encoding neural network (NN) and at least one training module. The final input image has been updated from the initial input image by a number of iterations of the iterative update. The iterative update includes generating a reconstructed image of the final input image based on the coded representation of the final input image by a decoding NN. One of a rate-distortion loss for the final input image or the number of iterations of the iterative update satisfies a pre-determined condition. An encoded image corresponding to the final input image is generated.

Content-adaptive online training with image substitution in neural image compression

An encoder comprising a processor configured to obtain candidate motion vectors (MVs) corresponding to neighboring blocks of a current block, the neighboring blocks neighboring the current block; obtain precisions of the candidate MVs; round the precisions to a target precision based on a rounding scheme; round the candidate MVs based on the target precision; perform pruning of the candidate MVs; generate a candidate list based on the rounding of the candidate MVs and the pruning; select one of the candidate MVs from the candidate list for encoding the current block; and encode an MV candidate index corresponding to the one of the candidate MVs that was selected in a bitstream.

Selection and signaling of motion vector (MV) precisions

There is included a method and apparatus comprising computer code configured to cause a processor or processors to perform obtaining a video bitstream, coding the video bitstream at least partly by a neural network, determining topology information and parameters of the neural network, signaling the determined topology information and the parameters of the neural network in a plurality of syntax elements associated with the coded video bitstream.

Techniques for signaling neural network topology and parameters in the coded video stream

A method of decoding an image with latent feature-domain intra-prediction is performed by at least one processor and includes receiving a set of latent blocks and for each of the blocks in the set of latent blocks: predicting a block, based on a set of previously recovered blocks; receiving a selection signal indicating a currently recovered block, based on the selection signal performing one of (1) and (2): (1) generating a compact residual, a set of residual context parameters, a decoded residual, and generating a first decoded block; (2) generating a second decoded block, based on a compact representation block and a set of context parameters. The method further includes generating a set of recovered blocks comprising each of the currently recovered blocks; generating a recovered latent image by merging all the blocks in the set of recovered blocks; and decoding the recovered latent image, to obtain a reconstructed image.

Neural image compression with latent feature-domain intra-prediction

A method, computer program, and computer system is provided for compressing a deep neural network model. Weight coefficients associated with a deep neural network are quantize and entropy-coded. The quantized and entropy-coded weight coefficients are locally smoothed. The smoothed weight coefficients are compressed based on applying a variational dropout to the weight coefficients.

Variational dropout with smoothness regularization for neural network model compression

A method of adaptive neural image compression with a hyperprior model by meta-learning is performed by at least one processor and includes generating a statistic feature, based on an input image and a hyperparameter, and generating a first shared feature and an estimated adaptive encoding parameter, encoding the input image to obtain a signal encoded image, based on the generated first shared feature and the generated estimated adaptive encoding parameter, generating a second shared feature and an estimated adaptive hyper encoding parameter, generating a hyper feature, based on the signal encoded image, the generated second shared feature, and the generated estimated adaptive hyper encoding parameter, and compressing the obtained signal encoded image, the generated statistic feature, and the generated hyper feature.

Method and apparatus for adaptive image compression with flexible hyperprior model by meta learning

Certain aspects of the present disclosure provide techniques for vectorization of transactions including: receiving electronic transaction information of one or more transactions of a user; for each transaction of the one or more transactions: segmenting the electronic transaction information of the transaction into one or more transaction words; generating a second transaction description related to the transaction; and identifying a category of the transaction; generating, based on the corresponding identified categories of the one or more transactions, a set of transaction history data of the user; providing the set of transaction history data of the user as an input to a machine learned model trained to output a set of word embedding vectors; determining, based on an output of the machine learned model comprising a set of word embedding vectors, a set of similar merchants; and providing the set of similar merchants for display to the user.

Vectorization of transactions

A method of feature substitution for end-to-end image compression, is performed by at least one processor and includes encoding an input image, using a first neural network, to generate an encoded representation, and quantizing the generated encoded representation, using a second neural network, to generate a compressed representation. The first neural network and the second neural network are trained by determining a rate loss, based on a bitrate of the generated compressed representation, and updating the generated encoded representation, based on the determined rate loss.

Method and apparatus for feature substitution for end-to-end image compression

Growing community of inventors

Cupertino, CA, United States of America

Wei W Wang