Online Adaptation in Image Coding with a 2-D Tarp Filter

Data Compression Conference |

Published by Institute of Electrical and Electronics Engineers, Inc.

On-line adaptation to nonstationary distributions is essential to good performance in image coding. Fixed-size contexts (with adaptive tables) are also widely used, in conjunction with arithmetic encoders, in state-of-the-art codecs. In contrast, we propose a simple two-dimensional filter that directly outputs the probability distribution function (PDF) estimate necessary to drive an adaptive arithmetic encoder. The filter is isotropic, in the sense that the impact of a previously encoded bit depends only on its value and distance to the bit to be coded. Surprisingly, this simple filter yields results comparable to or better than JPEG2000. It also brings an interesting distinction between on-line and off-line learning, and their relative importance in compression.