Amazon Technologies, Inc.

Yash Patel

R Manmatha

Srikar Appalaraju

Richard Threlkeld

Michael Paris

Server-specified subscription filters for long-lived client requests to fetch data in response to events. In one aspect, the techniques encompass a method performed by a set of one or more computing devices. The method includes the step of receiving a long-lived request to fetch data in response to events sent by a client computing device. The method further includes receiving a server-specified subscription filter for the long-lived request and executing the long-lived request. Executing the long-lived request includes creating a persistent function that uses the server-specified subscription filter to map a source event stream to a response event stream. The response event stream is provided to the client computing device. The server-specified subscription filter facilitates filtering of events fetched for the long-lived request in a way that may not be possible or impractical if the subscription client were required to specify the filter in the long-lived request.

Server-specified filters for long-lived client requests to fetch data in response to events

The present application relates to a multi-stage encoder/decoder system that provides image compression using hierarchical auto-regressive models and saliency-based masks. The multi-stage encoder/decoder system includes a first stage and a second stage of a trained image compression network, such that the second stage, based on the image compression performed by the first stage, identify certain redundancies that can be removed from the bit string to reduce the storage and bandwidth requirements. Additionally, by using saliency-based masks, distortions in different sections of the image can be weighted differently to further improve the image compression performance.

Hierarchical auto-regressive image compression system

Techniques for learned lossy image compression are described. A system may perform image compression using an image compression model that includes an encoder to compress an image and a decoder to reconstruct the image. The encoder and the decoder are trained using machine learning techniques. After training, the encoder can encode image data to generate compressed image data and the decoder can decode compressed image data to generate reconstructed image data.

Growing community of inventors

Palo Alto, CA, United States of America

Yash Patel