Studying Error Resilience Performance for a Feedback Channel Based Transform Domain Wyner-Ziv Video Codec

Abstract

Wyner-Ziv (WZ) video coding is an emerging video coding paradigm based on two major Information Theory results: the Slepian-Wolf and Wyner-Ziv theorems. One of the most interesting and used WZ video coding architectures makes use of a feedback channel (FC) to perform rate control at the decoder; in this context, the Slepian-Wolf coding module is typically based on turbo coding with puncturing. Because WZ coding is not based on the prediction loop used in conventional video coding but rather on a statistical approach where a decoder estimation of the frame to be coded is ‘corrected’ by the encoder, it provides intrinsic error resilience capabilities. This paper intends to study the error resilience performance of a feedback channel based transform domain WZ codec using appropriate scenarios and conditions, notably in comparison with the best performing H.264/AVC standard