Google Inc.

Semiconductor Components Industries, LLC

Nan Du

Andrew M Dai

Yonghui Wu

Kun Zhang

Yingwei Cui

Dengyong Zhou

Melvin Jose Johnson Premkumar

Izhak Shafran

Orhan Firat

Slav Petrov

Yanping Huang

Yuanzhong Xu

Aurko Roy

Maxim Krikun

David Richard So

Rohan Anil

David Wayne Jasinski

Xavier Eduardo Garcia

Gaurav Mishra

Linh Mai Tran

Yuan Xue

Dmitry Lepikhin

Yu-hui Chen

Zhen Xu

Siamak Shakeri

Yong Cheng

Jonathan Hudson Clark

Erica Ann Moreira

Yujing Zhang

Zachary Alexander Nado

Kefan Xiao

Jin Young Sohn

Eric Jun Jie Ni

Vladimir Feinberg

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training a neural network to perform any one or more of a variety of machine learning tasks. For example, the neural network can be configured as a generative neural network, e.g., an autoregressive generative neural network.

Training of large neural networks

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for extracting entities from conversation transcript data. One of the methods includes obtaining a conversation transcript sequence, processing the conversation transcript sequence using a span detection neural network configured to generate a set of text token spans; and for each text token span: processing a span representation using an entity name neural network to generate an entity name probability distribution over a set of entity names, each probability in the entity name probability distribution representing a likelihood that a corresponding entity name is a name of the entity referenced by the text token span; and processing the span representation using an entity status neural network to generate an entity status probability distribution over a set of entity statuses.

Learning to extract entities from conversations with neural networks

A deep state space generative model is augmented with intervention prediction. The state space model provides a principled way to capture the interactions among observations, interventions, critical event occurrences, true states, and associated uncertainty. The state space model can include a discrete-time hazard rate model that provides flexible fitting of general survival time distributions. The state space model can output a joint prediction of event risk, observation and intervention trajectories based on patterns in temporal progressions, and correlations between past measurements and interventions.

Machine-learned state space model for joint forecasting

Image sensors may include phase detection pixels that are used to determine the distance between the image sensor and objects in a scene. To account for variations in the angular response of phase detection pixels across a pixel array, calibration may be performed. During calibration, an image sensor may take images of targets at known distances from the sensor. The known distances may be used to determine scaling factors that account for differences between the initially calculated distances between the sensor and the targets and the actual distances between the sensor and the targets. The scaling factors may then be stored on the image sensor for future reference. During subsequent use of the image sensor, the image sensor may then obtain the appropriate scaling factor in real time for more accurate phase and distance determinations.

Image sensors with calibrated phase detection pixels

Growing community of inventors

San Jose, CA, United States of America

Nan Du