International organisation for standardisation organisation internationale de normalisation

3.2Output documents related to MPEG-7 Visual

No.	Title	TBP	Available
	15938-3 Visual
10566	Request for 15938-3:2000/Amd.4	N	09/04/24
10542	WD 1.0 of 15938-3/Amd.4 Video Signature Descriptors	N	09/05/08
10543	MPEG-7 Visual XM 35	N	09/05/15
10544	Description of Core Experiments in Video Signature Description development	N	09/04/24
10545	Disposition of Comments on ISO/IEC 15938-6:2003/PDAM 3	N	09/04/24
10546	Text of ISO/IEC 15938-6:2003/FPDAM 3 Reference Software for Image Signature Tools	N	09/05/08
10547	Disposition of Comments on ISO/IEC 15938-7:2003/PDAM 5	N	09/04/24
10548	Text of ISO/IEC 15938-7:2003/FPDAM 5 Conformance Testing for Image Signature Tools	N	09/06/12
10549	Disposition of Comments on ISO/IEC 15938-8:2002/PDAM 5	N	09/04/24
10550	Text of ISO/IEC 15938-8:2002/DAM 5 Extraction and Matching of Image Signature Tools	N	09/05/08

423002 MPEG-C Video Technologies

4.123001-4 and 23002-4 Reconfigurable Video Coding (RVC)

4.1.1General status of work

The two parts related to RVC (ISO/IEC 23001-4 Codec Configuration Representation in MPEG-B and ISO/IEC FCD 23002-4 Video Tool Library in MPEG-C) were progressed into FDIS status in Lausanne, but some final editing work was dedicated in Maui, in particular to guarantee a 1:1 matching between the textual description in the FDIS and the software implementation in Amd.1 which together with the conformance gives the “anchor” reference of normative FU input/output behavior. In particular, a number of inconsistencies were detected in the implementation of the AVC BP de-blocking filter, which needs to be split into two functional units to provide correct operation and timing.

Regarding conformance definitions in PDAM1, intense discussions were performed during the week and it was concluded that the currently defined conformance at the FU level is not sufficient to describe the normative behaviour, but interoperation of FUs, synchronization and timing is potentially even more relevant. It would however be too early to include any draft text on this in a Study of PDAM. Instead, a Core Experiment was started, which is expected to bring more clarification about viable solutions.
It is still the understanding that RVC is still in "phase 1", which is re-implementation of existing decoder conformance points

• 
  MPEG-4 SP and MPEG-4 AVC CBP in FDIS
  
• 
  MPEG-2 MP, MPEG-4 ASP, AVC HP and SVC to follow in Amd.2.
  

Due to ongoing investigation about best way of defining CABAC related parsers in the RVC/BSDL context, as well as due to the time that needed to be spent on the finalization of text and software/conformance issues, the progress on Amd.2 is slower than expected, such that issuing the PDAM will certainly be delayed by at least one more meeting cycle.
"Phase 2" could open up more options that were previously discussed in the RVC context, in particular

• 
  Possible simplification of standards development by adding new FUs. Whereas this approach sounds attractive, it appears clear that a real simplification (by adding / exchanging single FU modules) would only be given for cases where a preceding standard is fully implemented in RVC schema – it needs to be clarified whether RVC-based encoders are necessary in this context as well. First investigations on this were reported and will be continued in Core Experiments.
  
• 
  downloadable (on the fly) decoder solutions
  

More of this will be discussed and provided in the expected update of the vision document, which will be discussed in the AHG reflector.
The main activity (besides discussion of vision) until next meeting will be the conduction of Core Experiments related to

• 
  RVC Encoding Tools
  
• 
  RVC Conformance Testing
  (will be updated after 4 weeks)
  
• 
  Re-usability of tokens and parametrization
  
• 
  RVC-BSDL extension, in particular with relation to CABAC