This paper describes a prototype video coding platform meant for the conception and testing of multimedia products such as next-generation videophones. The platform is largely based on ITU-T Recommendation H.263, with a number of additional object-oriented quality enhancement features which make it especially well suited for very low bit-rate coding of "head-and-shoulders" video material typical of real-time multimedia applications, video teleconferencing, and video telephony. These features consist of: (1) segmentation into objects of interest, (2) segmentation-based prefiltering, (3) model-assisted rate control, (4) adaptive vector quantization, and finally (5) segmentation-based postfiltering. In the spirit of Recommendation H.263, these enhancements are modular and can be selectively turned on or off, thereby enabling a wide variety of coding modes.

We propose a postfiltering algorithm which adapts to global image quality as well as (optionally) to semantic image content extracted from the video sequence. This approach is in contrast to traditional postfiltering techniques which attempt to remove coding artifacts based on local signal characteristics only. Our postfilter is ideally suited to head-and-shoulders video coded at very low bit rates (less than 25.6 kbps), where coding artifacts are fairly strong and difficult to distinguish from fine image detail. Results are shown comparing head-and-shoulder sequences encoded at 16 kbps with an H.263-based codec to images postfiltered using the content-adaptive postfilter proposed. The postfilter manages to remove most of the mosquito artifacts introduced by the low-bit-rate coder while preserving a good rendition of facial detail.