Iterative generation of motion-compensated side information for distributed video coding

Distributed video coding (DVC) is a coding paradigm that gives the decoder the task to exploit the source statistics to achieve efficient compression. Current approaches to DVC rely on motion-compensated interpolation to generate at the decoder an estimation of the frame being decoded. This paper presents an iterative motion-compensated interpolation technique that takes advantage of all available information about the frame being estimated, not only the previous and posterior frames as is common practice. Simulation results show that the addition of this estimation technique to an existing DVC codec produces a 0.15 dB improvement in the PSNR of the rate-distortion plots.