Multiple codebook decoding of VQ compressed images

This paper proposes an implementation for vector quantizers (VQ) with one-codebook encoding and multiple-codebook decoding. The technique uses a convex projections (CP) based algorithm to iteratively project a coarsely encoded image onto a better codebook(s) during decoding. The objective of this approach is to encode edge vectors vestigially. This drastically reduces the number of edge vector representations at the encoder and hence results in fast searches. Also this method works better on images outside the training set since encoding is less dependent on the edges.