International organisation for standardisation organisation internationale de normalisation

Liaison activity

The JCT-3V did not send or receive formal liaison communications at this meeting.

12. Opening remarks

A new work item was brought into the JCT-3V at this meeting: multi-view frame compatible (MFC) extension of AVC.

It was noted that there was an urgent need for coordination with JCT-VC on HLS 3D/SHVC.

13. Contribution topic overview

The approximate subject categories and quantity of contributions per category for the meeting were summarized as follows.

• 
  AHG reports (section 2) (14)
  
• 
  CE1. View synthesis prediction and related (section 4.1) (22+6)
  
• 
  CE2: Disparity vector derivation and related (section 4.2) (5+16)
  
• 
  CE3: Inter-component prediction and related (section 4.3) (5+3)
  
• 
  CE4: Residual prediction and related (section 4.4) (5+3)
  
• 
  CE5: Inter-view / motion prediction and related (section 4.5) (10+7)
  
• 
  CE6: Depth intra coding and related (section 4.6) (11+19)
  
• 
  CE7: Coded depth representation (section 4.7) (3)
  
• 
  3DV Standards Development (section 5.1) (3+3+3+8)
  
• 
  High-Level Syntax (section 5.2) (4+14)
  
• 
  Other technical contributions on coding layer (section 5.3) (13+9)
  
• 
  Conformance testing (section 5.4) (0)
  
• 
  Software (section 5.5) (3)
  
• 
  Source Video Test Material (section 5.6) (1)
  
• 
  Alternative Depth Formats (section 5.7) (4)
  
• 
  Non-normative Contributions (section 5.8) (2)
  
• 
  Quality assessment (section 5.9) (2)
  

NOTE – The number of contributions noted in each category, as shown in parenthesis above, may not be 100% precise.

2. AHG reports

The activities of ad hoc groups that had been established at the prior meeting are discussed in this section.
4.2.1.1.1.1.1.1.1JCT3V-C0001 JCT-3V AHG Report: JCT-3V project management [J.-R. Ohm, G. J. Sullivan]

All documents from the preceding meeting had been made available at the document sites

The 14 ad hoc groups and core experiments had made progress, and various reports from those activities had been submitted.

The software for ATM versions 6 and 6.1 and HTM versions 5 and 5.1 had been prepared and released approximately as scheduled.

Preparations have been made to install "bug tracking" systems for software and text specifications. Such a system is expected to be extremely helpful to the software coordinators and text editors in JCT-3V and is planned to be taken into full operation after the 3^rd meeting. The bug tracker reports will be automatically forwarded to the group email reflector, where the issues could be further discussed.

A preliminary basis for the document subject allocation and meeting notes had been circulated to the participants.

4.2.1.1.1.1.1.1.2JCT3V-C0002 JCT-3V AHG report: MVC+D / 3D-AVC Draft and Test Model editing (AHG2) [M. M. Hannuksela, S. Hattori, Y. Chen, T. Suzuki, J.-R. Ohm, G. J. Sullivan]

Two deviations between software and text were noted.

The test model document was noted to not be fully up to date (some contributions missing as listed in the report).

Action item: Identify a separate editor for the test model.

It was suggested to contact contributors of technology personally, and remind them about their duties.

If not delivered in a reasonable time, we may possibly remove technology from spec.

4.2.1.1.1.1.1.1.3JCT3V-C0003 JCT-3V AHG Report: MV-HEVC / 3D-HEVC Test Model editing (AHG3) [G. Tech, K. Wegner, Y. Chen, T. Suzuki, S. Yea, J.-R. Ohm, G. J. Sullivan]

Deviation between SW and text was noted: SW based on HM6, text on HM8.

It was planned to split further versions of the 3D-HEVC test model document in two files: One containing the test model description and one containing working draft text of Annex G.

Migration into HM10 (text/software)? One option could be to start the migration with MV-HEVC related part first (until next meeting), and specific 3D tools afterwards. See further discussion under JCT3V-C0239.

4.2.1.1.1.1.1.1.4JCT3V-C0004 AHG4: 3D-AVC Software Integration (AHG4) [Dmytro Rusanovskyy, J. Y. Lee, J.Lin, D. Tian, O.Stankiewicz]

Software was delivered on time.

Inconsistency was found with adoption JCT3V-B0054 (was afterwards disabled in CTC).

Improvement compared to ATM 5: Approx. 2% total BR reduction (coded views).

4.2.1.1.1.1.1.1.5JCT3V-C0005 JCT-3V AHG Report: 3D-HEVC Software Integration (AHG5) [G. Tech, L. Zhang, Y. Chang, K. Wegner]

Software was delivered on time

Improvement compared to ATM 5: Approx. 1% total BR reduction (coded views).

Migration into HM10: see above under AHG3.

4.2.1.1.1.1.1.1.6JCT3V-C0006 JCT-3V AHG Report: 3D Coding Tool Testing (AHG6) [K. Mueller, A. Vetro]

No specific remarks were made about this report.

4.2.1.1.1.1.1.1.7JCT3V-C0007 JCT-3V AHG Report: 3D High level syntax (AHG7) [Y. Chen, T. Rusert]

No specific remarks were made about this report. Collaboration with JCT-VC experts (joint meeting and breakout) was planned for alignment of HL syntax between HSVC and MV-HEVC extensions.

4.2.1.1.1.1.1.1.8JCT3V-C0008 JCT-3V AHG Report: Alternative 3D Formats (AHG8) [M. Tanimoto, T. Senoh]

No specific remarks were made about this report.

For the warping format, coded results were produced which allow visual comparison against current 3D-HEVC (with conventional view synthesis) against the warping method. Expert viewing was prepared by N. Stefanofski.

4.2.1.1.1.1.1.1.9JCT3V-C0009 JCT-3V AHG Report: 3D Quality Assessment (AHG9) [T. Ebrahimi, A. Vetro, V. Baroncini]

Verification tests on MVC+D and 3D-AVC were discussed as possible future action item. For this test, the same data that could be used which are prepared for the planned collaboration with Qualinet on providing data for the investigation of quality metrics.

Potential verification test for MVC+D: Comparison of quality versus rate for 2-view + depth approach (used to generate 2 intermediate views) against 4-view MVC, with same range of baseline between outer views (which would be the coded views in MVC+D case).

4.2.1.1.1.1.1.1.10JCT3V-C0010 JCT-3V AHG Report: Inter Component Dependencies (AHG10) [J. Jung, J.-L. Lin]

No activity on assessing memory complexity; document JCT3V-C0170 could give a first step in that direction.

AHG report includes some more detailed analysis of various tools with inter-component dependencies.

For 3D-HEVC, deactivating the tools with inter-component dependency increases the bit rate by approx. 2.8% (for coded views).

Another contribution (JCT3V-C0168) not listed in the AHG report investigates impact of disabling texture-to-depth prediction.

4.2.1.1.1.1.1.1.11JCT3V-C0011 JCT-3V AHG Report: 3D Test Material (AHG11) [T. Wiegand, S. Yea]

New sequences offered by NICT: 4K-stereo, 3-camera+1-depth, CG content.

Partially not rectified, no camera data.

Viewing session to be prepared.

Another offer was reported for holoscopic material.

4.2.1.1.1.1.1.1.12JCT3V-C0012 JCT-3V AHG Report: Conformance testing development (AHG12) [T. Suzuki, D. Rusanovskyy]

No bitstreams available yet, partially due to problem with reference software (interlaced), but it is asserted to be possible to generate all bitstreams by the next meeting.

4.2.1.1.1.1.1.1.13JCT3V-C0013 JCT-3V AHG Report: DV constraint (AHG13) [Y.-W. Chen, I. Tomohiro]

It is suggested to introduce a maximum vertical disparity constraint of +/− 56 samples shift for 3D-HEVC. To be clarified: What are possible maximum disparity ranges in realistic non-rectified pictures? Would this be sufficient for UHD formats, or would the constraint need to be applied in a level dependent way? Would the applicability for applications such as robotics, automotive suffer by such a constraint?

4.2.1.1.1.1.1.1.14JCT3V-C0014 JCT-3V AHG Report: Reduced Resolution Depth Coding (AHG14) [K. Wegner, S. Shimizu]

One input on asymmetric video resolution (JCT3V-C0056) was identified. No input on comparing impact of full-resolution versus reduced-resolution depth (neither for 3D-HEVC nor 3D-AVC). Problem with software implementations, as in both software packages some tools exist which are not capable to operate with both full and reduced resolution.

3.