SIPI Video Compression Laboratory

The Video Compression Lab is devoted to theoretical issues and practical applications of image and video compression in emerging multimedia communications systems. The lab currently hosts two faculty members (Jay Kuo and Antonio Ortega), two post doctoral fellows, and around thirty graduate students, and is supported by several sources including companies such as Intel, HP, Rockwell, Texas Instruments, Lockheed Martin, Hughes, LG Electronics, Summitec, Intertop, Aegir Systems, ITRI and III. The research focuses on three major areas, namely, (i) novel compression techniques for multimedia data, including images, video, audion, graphics, etc. (ii) integration of image and video applications within heterogeneous and, possibly, unreliable network environments, and (iii) management of audio, image and video databases.

We are pursuing new methods, based on wavelets and efficient rate control/rate allocation techniques, for the coding of traditional media, such as images (we are actively involved in the JPEG 2000 standard process) and video. We are also exploring new techniques for coding of stereo images, graphics (2D and 3D) and audio. To tackle the lack of rate guarantees of typical networking environments (e.g. Internet) we are studying methods for transmission of video over lossy links (e.g. wireless links) and also considering efficient ways of transporting the video data, adapting to changes in the network state. Finally, our work on image and video databases focuses on developing effective techniques for content based retrieval, as well as in exploring resource allocation issues, such as those arising when we allow caching to be performed at intermediate nodes in the network.