An adaptive high-speed lossy data compression

An adaptive method for lossy data compression and the associated VLSI architecture have been developed. This scheme does not require a-priori knowledge of the source statistics and codebook training. The codebook is generated on the fly and is constantly updated to capture local textual features of data. The algorithm is proven to reach rate distortion function for memoryless sources. The authors also propose a computing architecture which consists of a vector quantizer and an encoded-data generator. By using this method, a high-speed VLSI processor with good local adaptivity, less complexity and fair compression ratio can be achieved.<>