Data Compression Laboratory
Welcome to the Data Compression Laboratory at the University of Washington's department of Electrical Engineering. We conduct research in the areas of image and video compression and image processing. The lab is run by Professor Eve Riskin and Professor Richard Ladner of Computer Science and Engineering. The laboratory is supported by the National Science Foundation, Microsoft Research, and Intel Corporation. |
| Research Information |
Current projects includes research into image compression and classification including our MobileASL project, near-lossless hyperspectral image compression, constant quality rate control in video coding and low bit-rate structured video coding. Past projects in our lab include research into vector quantization (VQ), application of VQ for edge detection and image browsing, VQ design for noisy channels, wavelets, image compression, halftoning, color palette management, unequal loss protection, layered video, and group testing algorithms for image compression. See our other pages for more information about the lab's graduate students, our publications, and courses. |
| Current Students |
|
| Publications |
Our recent publications are available via http. |
| Code |
We have C code available for full search VQ and tree search VQ from our html directory . These programs output codebook vectors that are close to the input data vectors in a mean square sense, but do not produce the intermediary indices. The file VQcode.tar.Z contains both sets of programs. We also have C code available for joint optimal wavelet packet best basis selection/bit allocation. In addition, Marko Slyz (mslyz@erim.org) has provided code to use with the TSVQ programs to produce codebook indices, or to decompress the indices to produce a reproduction of the data. The index sequence is somewhat robust (via resynchronization) to channel errors. The file indices_08_19_96.tar.Z contains these programs. |
| Useful Books and Lecture Notes |
Here is a selected bibliography of books useful to research in data compression. Professor Riskin has co-authored the textbook "Signals, Systems, and Transforms" with Phillips and Parr (Prentice Hall Publishers, 2003). There is also a set of notes to go with "Vector Quantization and Signal Compression" by Gersho and Gray (Kluwer Publishers, 1992). They are in postscript and can be used by anyone without charge. Please send any comments on the notes to Eve Riskin. |
| Collaborators |
|
| Alumni |
|
last updated 5/3/10 - rv