Network-driven Wyner-Ziv video coding using forward prediction

In some video applications such as video surveillance, a simple encoder is preferred with the computationally intensive tasks left to the decoder. The Wyner-Ziv Theorem showed that this goal is achievable by exploiting video source statistics at the decoder. In many existing Wyner-Ziv video coding schemes, many frames have to be intra. coded so that the decoder can derive sufficiently accurate side information from the frames. In this paper we present a new network-driven Wyner-Ziv method using forward prediction. The basic idea is to perform motion estimation at the decoder and send the motion information back to the encoder through a feedback channel. Experimental results show that our proposed approach can improve coding efficiency by up to 5dB.