eTech & Art
   New in 02!
  Reports from SIGGRAPH 2001

MPEG-4: The Next Generation for Interactive Media

Lecturers: Klaus Diepold, Radek Grzszczuk, Igor Pandzic, Eric Petajan, Iraj Sodagar, JC Spierer, and Gabriel Taubin

By Tai-San Choo

One of most eagerly awaited technologies in interactive media made its way to Alamo City this year in the rapidly advancing wave of MPEG-4 technology. MPEG-4 is the latest ISO/IEC standard developed by the Moving Pictures Experts Group, the committee responsible for widely used MPEG-1, MPEG-2, and MP3 formats that have revolutionized the interactive media industry.

This latest development takes into account the growing need for a standard media format across a variety of applications while fulfilling the needs of being in rich in quality, using efficient compression, and having fully interactive capabilities. Some of the more promising usages of MPEG-4 highlighted in this course were streaming-media delivery, videoconferencing, and real-time 2D and 3D animation.

With each successive lecture, each speaker gave an overview of the reasons for the development and adaptation of MPEG-4 as well as the variety of application value the format can bring. The original motivation for MPEG-4 came through the success of MPG-2 DVD format, in addition to the demand for highly interactive media.

The key mechanism of MPEG-4 technology is its ability to stream scenes through media objects (aural, video, audio-visual content that can be either natural or synthetic), allowing a separation of elements into layers. JC Spierer gave an overview of this object-oriented approach through its streaming and delivery. This technology significantly reduces the need for constant retransmission of non-essential and static elements of the media stream, and allows more crucial elements, like faces and central character body motion, to appear clearer in playback.

Some other benefits of MPEG-4 lie in the scalability of its video bit rate, as well as the ability to customize the degradation of video quality for the variety of end users. The bit rate is scalable from 20Kbps-6Mbps allowing faster transmission on lower bandwidths while still allowing high quality transmission on faster bandwidths.

The scalability of MPEG-4 allows for its usage in a multitude of devices, from HDTV, to basic web interactive multimedia, to video transmission on mobile devices. The technology has progresses far enough to be utilized in QuickTime v.6 as well as plugins for Real Player as the format becomes more widely used. Also, the speakers mentioned a boost in the need for videoconferencing technology following 9-11, and the resulting cut-backs in business flights.

Course organizer Klaus Diepold explained the process of object composition in MPEG4 by BIFS, Binary Format for Scene Description. BIFS inherits it basic structure from VRML, however VRML does not support 2D & 3D animation mixing or broadcast, while MPEG-4 has both capabilities.

Some other interesting areas of the course included the demonstrations on MPEG-4 real-time playback of face animation. Eric Petajan from face2face animation, inc. and Igon Pandzic from Visage Technologies AB went over their corresponding face animation technologies using morph targeting lip-synching and 3D mesh compression in conjuction with MPEG-4 to create, basically, 3D animated talking head models for a wide array of communication uses, like virtual call center agents, newscasters, storytellers, eLearning, multilingual communication, and entertainment.

Diepold ended the session with a demo of several MP4 media files. The level of interactivity was impressive and reminiscent of Flash media, allowing multiple layers of information and media objects to co-exist at the same time. For instance, during music videos lyrics to the songs scroll down in synch with the music. Also the video can be smoothly scaled up in size and back again without interrupting the stream, and also multiple video streams are capable of playing back at the same time.

When the MPEG-4 format was made standard there were hopes that its wide range of uses, scalability, rich quality, efficient compression, and highly interactive capabilities would allow it to become the most widely adapted multimedia format and it would boost the development of related standards. The lecturers of this course showed MPEG-4 is well on its way to reaching these goals and its future looks very promising.


MPEG Home Page
Web3D Consortium

MPEG Resources


This page is maintained by
Jan Hardenbergh
All photos you see in the 2002 reports are due to a generous loan of Cybershot digital cameras from SONY