Where for art thou MPEG-4?

It is an unfortunate fact of life that technological vision often precedes technological practicality by a long, long time. Who really invented the flying machine, Leonardo da Vinci? or the Wright brothers? And so it is with some of MPEG-4. For example, we had high hopes that object based coding could be used to represent real generic objects in video scenes. We will probably come close to a complete synthetic model of human facial features, perhaps even with a realistic texture mapping of real human faces. If this is combined with state-of-the-art textto-speech synthesizers, then a very low bit-rate representation of a talking person becomes possible, although it may not capture exactly an ‘original person appearing before a video camera. MPEG-4 will have the capability for ‘object-based’ scalability, i.e., audiovisual (AV) scenes will be a collection of separately encoded AV objects combined in a useful and pleasing manner appropriate to each multimedia system. Thus, a desktop viewer would see all objects in full resolution, whereas a mobile wireless viewer might see only a subset of the objects, and some of the visible objects may be portrayed with less than full spatial or temporal resolution. MPEG4 will continue MPEG-2’s foray into interactive applications, and here it should benefit from a strong collaboration with the ITU-T’s Low Bitrate Coding (LBC) group. Interactivity will evolve from simple VCR functionality toward a full panoply of networked multipoint communication Ill&GE COMMUNICATION