Generalized Multiple Description Coding with Correlating Transforms

Multiple description (MD) coding is source coding in which several descriptions of the source are produced such that various reconstruction qualities are obtained with different subsets of the descriptions. Unlike multiresolution or layered source coding, there is no hierarchy of descriptions; thus, MD coding is suitable for erasure channels or packet networks. A transform-based approach is developed for producing M descriptions of an N-tuple source. The descriptions are sets of transform coefficients and the transform coefficients of different descriptions are correlated so that missing coefficients can be estimated. This is a generalization of a method introduced by Orchard, Wang, Vaishampayay, and Reibman. Several transform optimization results are presented for memoryless Gaussian sources, including a complete solution of the (N=2, M=2) case with arbitrary weighting of the descriptions.