Content adaptive video coding and transport

Abstract

Summary form only given. Reducing the bandwidth needed to store or transport video signals has been an ongoing quest for some time now. The focus of existing video compression approaches has been to reduce the information redundancy in a video signal representation. Although today's compression rates would be unimaginable a decade ago, there still are many applications that can significantly benefit from even further improvements. Examples of these applications include wireless video and video-on-demand. In order to satisfy the stringent requirements of these applications, a fundamental change in this focus may be timely. Focusing on the "relevance" of information carried in a video signal in addition to the redundancy of the signal representation seems to be a promising direction for obtaining significant improvements in compression rates. We have used the content adaptive approach for resource allocation in compression and transport with promising results.