Amazon Technologies, Inc.

Other

Apple Inc.

Leland Stanford Junior University

Kwai Inc.

Beijing Transtreams Technology Co. Ltd.

Micron Technology Incorporated

The Ohio State University

Syracuse University

Insilixa, Inc.

Adv Antest Corporation

University of Ottawa

Yang Liu

Lingzhi Liu

Yongxiong Ren

Robert W Dutton

Roger T Howe

Pan Zheng

Dilek Zeynep Hakkani-Tur

Alice Marlin Brugger

Israel Pastrana Vicente

Steven Charles Michalske

Tracey Adam Finger

Daniel Satoru Hamaguchi

Di Jin

Ping Lu

Bedrich Mosinger

Ken May

Heng Liu

Jiangli Zhu

Tingjun Xie

Qiang Li

Arjang Hassibi

Arindam Mandal

Juane Li

Kah Kuen Fu

Arun Manickam

Seokhwan Kim

Biao Chen

Tagyoung Chung

Premkumar Natarajan

Kshama Jirage

Andrea Cuppoletti

Jing Huang

Behnam Hedayatnia

Mingming Lu

Xiaorui Wang

Aaron Lee

Scott A Myers

Richard Hung Minh Dinh

Cheng Kai Chen

Robinson Piramuthu

Benjamin J Pope

Gokhan Tur

Christopher R Wilk

Ashutosh Yugesh Shukla

Ronald W Davis

Seungjune Jeon

James G Droppo

Rudolph Van Der Merwe

Scott P Porter

Ruchir M Dave

Chao Wang

Saurabh Gupta

Wei Wang

Ted J Cooper

Tyler B Cater

Zhaoyin D Wu

Ieyuki Kawashima

Geoffrey Zhang

Zhen Peng

Xin Chen

Charles See Yeung Kwong

Jaron Waldman

Ming Sun

Eric N Nyland

Kevin M Froese

Alan C Wong

Jan Erik Solem

Ihtesham H Chowdhury

Yu Xu

Samuel Noble

David A Hurrell

Anna Chen Santos

Viktor Rozgic

Anthony P Grazian

David MacNeil

Xiaqing Wu

Hernan R Eguiluz

Hanlee P Ji

Qingming Tang

John Kassebaum

Vincent Tabard-Cossa

Eric D Schlakman

Salome Bavetta

Teemu P Sipila

Billy Tsz Cheong Lau

Miguel S Sanchez-Sandoval

Tianyang Ma

Xinguo Zhang

Ameya A Mhatre

Benjamin M Russo

Ze'ev Raz

Sungjin Lee

Sanqiang Zhao

Qin Zhang

Eshan Bhatnagar

Yelin Kim

Thomas Irvine Nelson

Ronald L Jett

Andrew M Fallon

Yen Lai

A dialog management system that coordinates system dialog responses based on natural language guidelines which provide non-deterministic ways for the system to properly respond to a dialog input based on the dialog history/context. For each input, an appropriate guideline is selected by a machine learning component based on the dialog history. The guideline is then sent, along with the dialog history, to a downstream machine learning component to determine an appropriate dialog system response.

Virtual assistant dialog management

A processing device in a memory sub-system traverses a plurality of management units of a memory device at a defined scan/read refresh frequency. For every management unit of the plurality of management units, the processing device identifies a page satisfying a lowest sensing overhead criterion, and senses data of the identified page without transferring the data out of the memory device. A non-transitory computer readable medium includes program instructions that when executed by a processing device, cause the processing device to perform operations of traversing a plurality of management units of a memory device at a defined scan/read refresh frequency. For every management unit, the processing device identifies a page satisfying a lowest sensing overhead criterion, and senses data of the identified page without transferring the data out of the memory device.

Efficient periodic backend refresh reads for reducing bit error rate in memory devices

In accordance with one disclosed method, first data representing first text may be processed using a first machine learning model configured to generate second data representing an explanation as to why the first text is humorous. The second data may be processed, together with the first data, by a second machine learning model configured to generate a value indicating that the first data and the second data correspond to a pun. A device may be caused to take at least a first action based at least in part on the value.

Virtual assistant humor management

Methods and apparatuses for automatic speech recognition are provided. The method includes: generating a weight matrix for a layer of a plurality of layers in a neural network; dividing the weight matrix into a plurality of blocks, each block including a plurality of weights; selecting a set of blocks from the plurality of blocks for block-wise pruning by minimizing a cost function subject to a pre-determined block-wise constraint; and generating a block-wise pruned weight matrix by setting one or more weights in the set of blocks to zero. The weight matrix includes a set of weights associated with the layer, the plurality of layers includes a first layer receiving a first input associated with one or more audio feature sequences, and the plurality of layers are executed on one or more processors.

Methods and devices for structured pruning for automatic speech recognition

Systems and techniques are described for suppressing false positive notifications for detected objects. In various examples, first bounding box data indicating a detection of a first class of object in a first frame of image data may be received from an object detector. Second bounding box data indicating a prior detection of the first class of object in a second frame of image data may be determined. A first value representing a similarity between the first bounding box data and the second bounding box data may be determined. A notification associated with the detection of the first class of object in the first frame of image data may be suppressed based at least in part on the first value.

Filtering false positive computer-vision-based object detection events

Methods and apparatuses are provided for serving multiple models with a single engine. The method includes: providing a single engine for a plurality of models on a server, where the server includes at least one graphics processing unit (GPU) and memory coupled with the at least one GPU; loading the plurality of models onto the memory of the server at once; and serving, by the single engine, the plurality of models, where the single engine accommodates all structures and weights of the plurality of models with a shared input and output of the memory.

Methods and devices for multiple model single deployment engine implementation

In accordance with one disclosed method, first data representing one or more first keywords may be processed by at least one first component to determine a first pair of words suitable for generating a pun. The first pair of words may then be processed using a first machine learning model configured to generate an output pun based at least in part on one or more input keywords and an input word pair such that the output pun is contextually related to the one or more input keywords. A first pun that is contextually related to the one or more first keywords may be received from the first machine learning model, and a first device may be caused to output a representation of the first pun.

Techniques for generating a natural language response to a user input of a dialog are described. A system receives a natural language user input of a dialog and determines dialog history data including a previous natural language user input of the dialog. Based on the first natural language user input and the dialog history data, the system generates at least a first question associated with the natural language user input. Based on the first natural language user input and the dialog history data, the system generates at least a first answer to the at least first question. Using the dialog history data, the first natural language question, and the first natural language answer, the system generates an output responsive to the natural language user input.

Natural language response generation

The present disclosure provides methods and compositions for surface functionalization of solid substrates. The compositions include functionalized silanes and nucleic acid constructs which may react to immobilize the nucleic acid constructs on the surface on the solid substrate. The disclosure also provides methods for immobilization of silanes and nucleic acid constructs on the surface of the substrate.

Methods and compositions for surface functionalization of optical semiconductor-integrated biochips

A processing device in a memory sub-system performs a first media scan operation with respect to a plurality of memory pages addressable by the ordinary wordline, wherein each page of the plurality of memory pages is contained by a respective management unit, and responsive to determining that a value of a data state metric of a memory page of the plurality of memory page addressable by the ordinary wordline satisfies a specified condition, performs a first media management operation with respect to a management unit containing the memory page. The processing device further performs a second media scan operation with respect to a plurality of memory pages addressable by the mandatory wordline, wherein each page of the plurality of memory pages is contained by the respective management unit, and responsive to determining that a value of the data state metric of a memory page of the plurality of memory page addressable by the mandatory wordline satisfies the specified condition, performs a second media management operation with respect to the management unit containing the memory page.

Reducing bit error rate in memory devices

Techniques for generating opinion-based content responsive to a user input are described. The system may receive a user input, and determine dialog context data corresponding to a dialog between a user and the system, and including the user input. The system may determine generation of content responsive to the user input requires opinion-based knowledge, and may extract entities from the dialog context data, and determine natural language data of a knowledge base that includes entities similar to the extracted entities. The system may processes the natural language data and the dialog context data to determine a subset of the natural language data that is responsive to the user input. The system may generate output data responsive to the user input using the responsive natural language data and the dialog context.

Opinion-based natural language response generation

An aerial vehicle is configured to operate within indoor spaces. The aerial vehicle is programmed with positions of waypoints within three-dimensional space, and to calculate a trajectory for traveling through such waypoints in a manner that minimizes snap of the aerial vehicle. Where a distance between a pair of the waypoints is sufficiently long, the aerial vehicle inserts intervening waypoints for planning purposes, and programs the aerial vehicle to travel at a maximum speed between the intervening waypoints. Upon detecting an obstacle with a first range using one or more sensors, the aerial vehicle reduces its speed and monitors a second, shorter range using the sensors, and compensates for motion between such readings.

Indoor navigation and obstacle avoidance for unmanned aerial vehicles

An aerial vehicle is configured to calculate ranges to objects around the aerial vehicle when operating within indoor spaces, using a LIDAR sensor or another range sensor. The aerial vehicle calculates ranges within a plurality of sectors around the aerial vehicle and identifies a minimum distance measurement for each of the sectors. Sets of adjacent sectors having distance measurements above a threshold are identified, and bearings and minimum distance measurements of the sets of adjacent sectors are determined. When the aerial vehicle detects an object within a flight path, the aerial vehicle selects one of the sets of adjacent sectors based on the minimum distance measurements and executes a braking maneuver in a direction of the selected one of the sets of adjacent sectors.

Selecting paths for indoor obstacle avoidance by unmanned aerial vehicles

Techniques for rendering visual content, in response to one or more utterances, are described. A device receives one or more utterances that define a parameter(s) for desired output content. A system (or the device) identifies natural language data corresponding to the desired content, and uses natural language generation processes to update the natural language data based on the parameter(s). The system (or the device) then generates an image based on the updated natural language data. The system (or the device) also generates video data of an avatar. The device displays the image and the avatar, and synchronizes movements of the avatar with output of synthesized speech of the updated natural language data. The device may also display subtitles of the updated natural language data, and cause a word of the subtitles to be emphasized when synthesized speech of the word is being output.

Virtual conversational companion

A system may be configured to detect custom acoustic events, where the system generates an acoustic event profile for the custom acoustic event based on a natural language description provided by a user and using an audio sample of the described acoustic event. For example, the user may describe the custom acoustic event as 'dog bark.' The system may ask the user questions to refine the description (e.g., dog breed, dog gender, age, etc.). Using an audio sample of the refined description, the system may then determine that audio captured in the user's environment is a potential sample of the custom acoustic event. Such captured audio may be presented to the user for confirmation, and then may be used to detect future occurrences of the custom acoustic event in the user's environment.

Acoustic event detection

Growing community of inventors

San Jose, CA, United States of America

Yang Liu