Affirmed Networks, Inc.

Microsoft Technology Licensing, LLC

Paul Alexander

John C Sievers

Srinivasan Venkatraman

Piotr Galecki

Thuy Fernandes

John Sievers

Systems and methods of transcoding video sequences for delivery in a mobile network. A source video sequence including source video frames is received and a source video frame type associated with each of the source video frames is determined based on a differential between a composition time stamp (CTS) and a decoding time stamp (DTS) associated with each of the source video frames. A frame type ratio value for the target frame-types is calculated based on a longest mini-GOP length. A target frame size of each target video frame is determined based on the frame type ratio value and a target transcoding rate, and each of the target video frames in a target video track is encoded to the target frame size to form a target video sequence.

Transcoding using time stamps

Systems and methods are described herein for adaptively transcoding media in a communications network such that computer processing resources are effectively utilized to deal with varying traffic patterns in the communications network. A computing device receives a request to transcode content and determines a first amount of tokens based on a number of CPU cores associated with the computing device, and a currently used amount of tokens. The computing device determines a second amount of tokens associated with transcoding the content and when the first amount of tokens is less than the second amount of tokens, the computing device bypasses transcoding for the content or adjusts a transcoding parameter to reduce a number of tokens associated with transcoding the content.

Adaptive content optimization

Growing community of inventors

Lexington, MA, United States of America

Paul Alexander