Meta Platforms Technologies, LLC

Meta Platforms, Inc.

Novumind Limited

Yilei Li

Sapna A Shroff

Sebastian Sztuk

Pingfan Meng

Jun Hu

Jiangtao Kuang

Emily Wang

Meng Li

Amy Lawson Bearman

Cheng-Ying Ko

Alireza Dirafzoon

Zhaonian Zhang

Ruchir Srivastava

Miao Li

Danni Luo

Jeffrey Weifung Lien

David Tao

Demetrios Basil Karanikos

Linsen Bie

In one embodiment, a method includes accessing from a client system associated with a first user sensor signals captured by sensors of the client system, wherein the client system comprises a plurality of sensors, and wherein the sensors signals are accessed from the sensors based on cascading model policies, wherein each cascading model policy utilizes one or more of a respective cost or relevance associated with each sensor, detecting a change in a context of the first user associated with an activity of the first user based on machine-learning models and the sensor signals, wherein the change in the context of the first user satisfies a trigger condition associated with the activity, and responsive to the detected change in the context of the first user automatically capturing visual data by cameras of the client system.

Auto-capture of interesting moments by assistant systems

A method for using cameras in an augmented reality headset is provided. The method includes receiving a signal from a sensor mounted on a headset worn by a user, the signal being indicative of a user intention for capturing an image. The method also includes identifying the user intention for capturing the image, based on a model to classify the signal from the sensor according to the user intention, selecting a first image capturing device in the headset based on a specification of the first image capturing device and the user intention for capturing the image, and capturing the image with the first image capturing device. An augmented reality headset, a memory storing instructions, and a processor to execute the instructions to cause the augmented reality headset as above are also provided.

Camera control using system sensor data

Disclosed herein includes a system, a method, and a device for improving power efficiency of a neural network implemented in an AI chip. In a neural network, large amounts of computations for multiply and accumulate can result in frequent toggles or transitions in states of logic circuits in the AI chip. Such frequent toggles or transitions of states of logic circuits can cause a large overall power consumption. In one aspect, to minimize the number of toggles, a sequence or order of computations can be rearranged. In one approach, total hamming distances for weights or input strings in different arrangements or sequences can be identified, and an arrangement or a sequence of weights or input strings with a reduced or minimum total hamming distance can be identified. An arrangement or a sequence of weights that render a reduced total hamming distance can be identified.

Neural network based on total hamming distance

A multiplier for calculating a multiplication of a first fixed point number and a second fixed point number comprises a converter and a restoration circuit. The converter is configured to convert the first fixed point number to a sign, a mantissa, and an exponent. At least one of a bit width of the sign, a bit width of the mantissa, and a bit width of the exponent is dynamically configured based on a position of a layer associated with the first fixed point number in a neural network, a position of a pixel in an input feature map associated with the first fixed point number, and/or a channel associated with the first fixed point number. The restoration circuit is configured to calculate the multiplication based on the sign, the mantissa, the exponent, and the second fixed point number.

Method and system for elastic precision enhancement using dynamic shifting in neural networks

Growing community of inventors

Fremont, CA, United States of America

Yilei Li