Multiple representation coding for error-resilient video delivery

The quality of the video delivered to the end-users is severely hampered by packet losses over the network, especially in wireless and mobile environments. In this paper, we address the challenge of error-resilient video delivery using a scheme we term as Multiple Representation Coding (MRC). In the MRC scheme, multiple independently decodable representations are generated from the source video. These multiple representations are then transmitted as a single video stream using a novel GOP interleaving (GOP: Group of Pictures) scheme. The proposed GOP interleaver ensures that spatio-temporally colocated segments of the video sequence corresponding to different multiple representations are not impaired by the same burst loss. The results demonstrate that when the transmitted bitstream is impaired by a burst loss spanning multiple frames, the MRC scheme can give a PSNR gain on the order of 2-6 dBs over the conventional full-size encoding and transmission of the video.