A progressive universal noiseless coder

Abstract

Abstruct-We combine pruned tree-structured vector quantization (pruned TSVQ) with Itoh's universal noiseless coder. By combining pruned TSVQ with universal noiseless coding, we benefit from the "successive approximation" capabilities of TSVQ, thereby allowing progressive transmission of images, while retaining the ability to noiselessly encode images of unknown statistics in a provably asymptotically optimal fashion. Noiseless compression results are comparable to Ziv-Lempel and arithmetic coding for both images and finely quantized Gaussian sources.

Strong Successive Refinability and Rate-Distortion-Complexity Tradeoff

“…Note that the tree structured vector quantization (TVSQ) has been extensively studied, and also has a successive approximation property. For example, in [21], Effros et al combined pruned TVSQ with a universal noiseless coder which enables progressive transmission of sources. While this approach guarantees optimality at zero distortion, it cannot achieve the rate-distortion function in general.…”