The Joint Collaborative Team on Video Coding (jct-vc) of itu-t wp3/16 and iso/iec jtc 1/ sc 29/ wg 11 held its tenth meeting during 11-20 July 2012 at the City Conference Centre (ccc), a k. a

JCTVC-J0003 JCT-VC AHG report: Software development and HM software technical evaluation (AHG3) [F. Bossen, D. Flynn, K. Sühring]

A brief summary of activities related to each mandate is given below.

1. Development of the software was coordinated with the parties needing to integrate changes. A single track of development was pursued. The distribution of the software was made available through the SVN servers set up at HHI and the BBC, as announced on the JCT-VC email reflector.

2. Version 7.0 of the software was delivered to schedule and reference configuration encodings were provided according to the common test conditions through an ftp site at the BBC.

ftp://ftp.kw.bbc.co.uk/hevc/hm-7.0-anchors/

3. Version 7.1 of the software was delivered ahead of the 10th JCT-VC meeting.

4. Some high-level adoptions were still outstanding at the time of writing.

Multiple versions of the HM software had been produced and announced on the JCT-VC email reflector. The changes made for each version were summarized in the report. A detailed history of all changes made to the software can be viewed at http://hevc.kw.bbc.co.uk/trac/timeline.

Released versions of the software are available on the SVN server at the following URL:

https://hevc.hhi.fraunhofer.de/svn/svn_HEVCSoftware/tags/version_number,

where version_number corresponds to one of the versions described below (eg., HM-7.0). Intermediate code submissions can be found on a variety of branches available at:

https://hevc.hhi.fraunhofer.de/svn/svn_HEVCSoftware/branches/branch_name,

where branch_name corresponds to a branch (eg., HM-7.0-dev).

Version 7.0 of the software was released on 23rd May 2012. It includes all the changes adopted at the 9th JCT-VC meeting that affect the common test conditions. This release was announced on the email reflector.

The performance change since HM-6.0 was tabulated in the report.
Some gain was achieved from adding AMP, but a number of small losses tended to offset that. Class F AI HE10 showed substantial gain, primarily due to adding the transform skip feature.
Version 7.1 of the software was released on 30th June 2012. It contains a number of bug fixes and the majority of adoptions form the 9th JCT-VC meeting that do not affect the common conditions. A number of integrations were still outstanding at the point of writing. There is virtually no performance change between HM-7.0 and HM-7.1 under the common conditions.

Version 7.2 was planned for release during the current meeting.

In addition to the regular HM development process, one branch was created to expose tools to a wider audience:

• 
  HM-7.1-dev-ahg13, which contains contains modifications to the reference picture buffers and list construction.
  

Recommendations of the AHG were as follows:

• 
  Continue to develop reference software based on HM version 7.1 and improve its quality.
  
• 
  Remove macros introduced inHM previous HMversions before startingintegration towards HM8.0 such as to make the software more readable
  
• 
  Continue to identify bugs and discrepancies with text, and address them
  
• 
  Test reference software more extensively outside of common test conditions
  

Several simplifications (restriction of motion comp in small PUs, SAO, etc.) added up in losses of up to 1%; AMP compensates for that such that the performance of HM7 is only very slightly inferior than HM6. The approx. 1% loss appears in HE10, where AMP was already switched on before (except class F, where transform skip which was adopted provides additional gain (4% rate reduction of HM7 vs. HM6).

JCTVC-J0004 JCT-VC AHG report: High-level parallelism (AHG4) [M. Horowitz (eBrisk), M. Coban (Qualcomm), F. Henry (Orange Labs), K. Kazui (Fujitsu), A. Segall (Sharp Labs), W. Wan (Broadcom), S. Wenger (Vidyo), M. Zhou (TI)]

There was no significant email activity for this AHG.

The chairs of AHG4 were unaware of any high-level parallel processing related open issues at this time.

The relevant changes made to the HM were reviewed.

The 22 related input documents to the Stockholm meeting were listed and categorized into five categories:

• 
  General (not specific to tiles or WPP)
  
• 
  Tiles
  
• 
  WPP (wavefront parallel processing)
  
• 
  Entry points
  
• 
  P&L (profile and level)