INTERNATIONAL ORGANISATION FOR STANDARDISATION

ORGANISATION INTERNATIONALE DE NORMALISATION

ISO/IEC JTC1/SC29/WG11

CODING OF MOVING PICTURES AND AUDIO

ISO-IEC/JTC1/SC29/WG11N1355

MPEG96

Sep-Oct 1996 / Chicago

Source: Convener of mpeg (iso-iec/jtc1/sc29/wg11)
Status: Approved by WG11
Subject: mpeg Press Release
Date: 2 October 1996

 

Emmy awarded for the development of MPEG standards

[The following is quoted from the press notice released by the American National Standards Institute, ANSI]

NEW YORK, October 2, 1996--The National Academy of Television Arts and Sciences (NATAS) awarded its 1995-1996 Engineering Emmy for Outstanding Achievement in Technological Development to the International Organization for Standardization (ISO) and the International Electrotechnical Commission (IEC) for their standardization work in media, the American National Standards Institute (ANSI), the U.S. member body of ISO announced today.

The award was presented to a joint ISO/IEC international subcommittee that developed standards (MPEG-1, MPEG-2, and JPEG) related to coded representation of video and systems for digital compression and coding of continuous tone still images. MPEG compression technology is a cost-effective means of delivering high quality audio and video programming to consumers by reducing the bandwidth necessary to carry the signal to homes and businesses across the globe. JPEG has revolutionized still image technology for both consumers and commercial interests with cost effective high quality still pictures.

Hundreds of dedicated individuals from 26 countries worked to develop each of these standards, and this award represents an excellent achievement by the international community, said Dr. Daniel T. Lee of Hewlett-Packard Company and leader of the working group, responsible for developing the JPEG standard.

The standards being recognized are: MPEG 1 (ISO/IEC 11172, Coding of moving pictures and associated audio for digital storage media at up to about 1,5 Mbit/s); MPEG-2 (ISO/IEC 13818, Generic coding of moving pictures and associated audio information); and JPEG (ISO/IEC 10918, Digital compression and coding of continuous-tone still images). The standards were developed by a subcommittee (Coding of audio, picture, multimedia, and hypermedia information) of the ISO/IEC Joint Technical Committee 1 on information technology. JTC-1 is administered by ANSI on behalf of the United States.

The Emmy Award was presented October 1 at a ceremony held in New York City. Accepting the award on behalf of the ISO/IEC Subcommittee were: Dr. Hiroshi Yasuda, Chairman of Subcommittee 29, Japan; Dr. Leonardo Chiariglione, Convener, Subcommittee 29, Working Group 11, Italy; Dr. Daniel Lee, Convener, Subcommittee 29, Working Group 1, United States; and Ms. Narumi Hirose, Secretariat, Subcommittee 29, Japan.

[End Quote ANSI press release]

Dr. Leonardo Chiariglione, Convener of MPEG, said: "The Emmy Award received 1st October by MPEG for its achievements in video compression is a much deserved, if partial, recognition of the importance of the MPEG-1 and MPEG-2 standards. The hundreds and hundreds of experts who, over the years, have produced these epoch-making pieces of technology, have created an unrivaled model of technical cooperation for everybody working in IT-standardization".

The Emmy award was presented at the same time that MPEG held its 36th meeting in Chicago, Illinois (USA). The meeting took place from 30 September - 2 October 1996, and was hosted by the United States National Standards Body (USNB) and organized by Motorola.

MPEG has taken on board a new Work Item, called 'Audiovisual Content Description Interface'. The idea is to add, to the coded representation of audiovisual objects, information about texture, shape, color (video), frequency range (audio), and also higher level semantic information, such as 'beach', or 'flute', or 'male voice', or 'car', 'blue'. With this information, audiovisual information can be found and retrieved in a standardized way. Finding information, e.g. on the WWW, is getting increasingly important. Currently, only text-based content can be well indexed and searched for; there is an obvious need to for the same possibilities for audiovisual content. The preliminary workplan indicates the standard will be ready in 1999. This new standard will be called MPEG-7.

Details on the 36th MPEG meeting

A detailed list of the results of the 36th meeting follows below. The list is largely organized according to the several sub groups of MPEG.

Requirements

The re-instated requirements group has begun an assessment of what is needed in the MPEG-4 standard for the support of 'multimedia database storage and retrieval'. A profile is being drafted for these applications, that need scalable coding, searching for specific objects in a scene, and browsing through - having access to - these individual objects. MPEG members from the audiovisual industry will bring further requirements for this profile to MPEG. Because many people still think that MPEG-4 is about 'low bitrates', it is very important to note that this profile will require MPEG­4's new capabilities while coding at a high quality - and therefore high bitrates. Currently, there is one other profile defined, for real-time audio-visual communication. It is expected that more profiles will be defined in the near future.

Systems

The systems group focused on the preparation of the Working Draft, to be ready in November. A first verification model based on the current specification has been written. It allows basic MPEG-4 functionalities such as composition and content based interaction.

The architecture developed by Systems was adopted for the integration of natural and synthetic material, but the inclusion of synthetic material, notably 3D objects, poses new requirements that need further study. The final system architecture should include comparable functions achieved by existing systems such as VRML 2.0. Another difficult topic is developing a signaling systems that can be used in interactive, conversational and broadcast services. The systems group looks at what was done in MPEG-2 (the so-called DSM-CC, Digital Storage Media - Command & Control), and also the signaling used by the new ITU-T standard for low bitrate videotelephony, H.245.

Integration of synthetic audio and video material

A verification model has been drafted for body and face animation, allowing the seamless integration of 'animated people' in natural scenes. As also text-to-speech systems are addressed, and the interface to the synthetic head is specified, these artificial people will be able to generate speech from written text. Currently, the text-to-speech is limited to languages from the Far East: Japanese and Korean. Other languages will follow. The standardization only addresses the interface to these text-to-speech systems, not the systems themselves.

Other items for standardization that are in the verification model are text and graphics overlays, as these are very useful visual objects in multimedia applications.

Audio

The audio group produced a preliminary version of the Working Draft for MPEG-4 Audio, containing detailed technical descriptions of all the tools comprising the VM-2.0. Currently, the VM supports the following functionalities: compression, pitch scaling and time scaling. The compression functionality is flexible in terms of bandwidth, bit-rate and complexity.

Video

In the MPEG-4 video group, numerous technical contributions have been reviewed, and the most promising proposals have been accepted as a part of the standard. These new elements of the MPEG-4 standard improve considerably the compression efficiency in intra frames. Also, the resilience against transmission errors is drastically improved with the introduction of resynchronization words. In co­operation with the implementation study group, the computational complexity of encoding and decoding video objects with an arbitrary shape has been reduced significantly.

For the coming weeks, over 20 so-called 'core experiments' have been planned. In the MPEG-meeting of November in Brazil, a number of promising core experiments are expected to again improve the standard considerably. Notably, the coding of the shape of objects will be changed, to obtain a much more efficient scheme. That meeting will also fix the first Working Draft of MPEG-4 Video, giving the first draft of the final standard.

The extension of the MPEG-2 video standard for multi-view applications (e.g. used for stereoscopic video) has been promoted to a final International Standard. It foresees higher compression of the right view of stereoscopic video by exploiting the similarity between the left and the right view.

Test activities

In the last two years the test group has coordinated several testing activities aimed at verifying the performance of the MPEG-2 video coding standard at different profiles and levels. At this meeting the results of tests carried out on stereoscopic sequences coded with the MPEG-2 multiview profile (ISO/IEC 13818-2/AM3) were presented and discussed.

These tests were carried out at three different test sites located in Japan, Germany and Canada. The results of the different test sites are consistent with each other and show that in general, at the tested bitrates, the observers judged that the MPEG-2 multiview profile coding scheme did not introduce annoying coding artifacts.

Next to addressing MPEG-2 issues, the test subgroup has continued the work to define test methods and procedures for the July '97 MPEG-4 tests. These tests will be aimed at checking the status of the Verification Models of Audio and Video against the requirements for the standard, and against technology available in the market place, as well as newly emerging technology.

The verification tests mentioned above are meant to check the performance of specific implementations of the complete standard (including the coding of natural and synthetic audio and video and the systems layer) This is a challenging goal, that requires, in a very early stage, the availability of fast software decoders or even an MPEG­4 system implemented in hardware.

Implementation Studies

Using a performance orientated implementation of the video verification model (VM) produced by members of the implementation studies group, a tenfold increase in performance has been achieved against the existing software implementations of the VM. Detailed profiling of this software identified the key performance critical components of the standard. Based upon this information, the video and implementation studies groups have jointly started an activity to reduce the complexity of the identified modules. There is still considerable scope to further improve the performance of the software and then in the future to include platform-dependent optimizations.

Another key activity performed by the implementation studies group was the identification of ways to gracefully degrade the computational complexity under conditions of high processing demands from either MPEG-4 itself or other co-existent applications, for instance when the decoder runs in software on a personal computer. These techniques, if proven feasible, will lead to increased service availability to the user.

Background

Over 300 experts took part in the work in Chicago, working in several sub groups on specific issues. The next meeting will take place from 18 through 22 November in Maceiņ, Brazil, hosted by ABNT (the Brazilian national standards body).

For further information, please contact:

Dr. Leonardo Chiariglione, (Convenor of MPEG)

CSELT

Via G. Reiss Romoli, 274

10148 Torino, ITALY

Tel.: +39 11 228 6120; Fax: +39 11 228 6299

Email: leonardo.chiariglione@cselt.it

or refer to the MPEG homepage:

http://www.cselt.it/mpeg

and to other pages containing information on MPEG-4 elements:

http://www-elec.enst.fr/msdl

http://www.es.com/mpeg4-snhc