Multiple description coding by successive quantization

Two structured multiple description (MD) vector quantization schemes with an iterative technique for designing the codebooks and partitions are proposed. The schemes are derived from the recent theoretical work by Chen et al. In the first scheme, the central decoder is formed by the weighted sum of the side codebooks, whereas the second scheme employs the optimum central decoder. The objective of the proposed iterative method is to minimize a Lagrangian cost function (defined as the weighted sum of the central and side distortions) to jointly design the side codebooks and find the associated partitions. The optimal parameters for minimizing the central distortion are also found. Simulations demonstrate that the proposed methods achieve performance close to that of the unstructured, full-search MD quantizer with considerably less complexity and with only a few iterations.