Universal Coding of Function Spaces as a Model for Signal Compression

We address the problem of signal compression, basing on the mathematical model, in which a set of all possible signals is considered as a function space with a metric . The main attention is focused on the minimization of the size of compressed representation, when function characteristics are not known precisely. Keywords. data compression, universal coding, epsilon- entropy, Kolmogorov entropy.

• Type: Conference
• Year: 2004
• Where: DCC
• Authors: Boris Ryabko, Jaakko Astola
• Keywords: Computer Graphics, DCC 2004, Function Characteristics, Kolmogorov Entropy, Signal Compression