International organisation for standardisation organisation internationale de normalisation



Yüklə 8,63 Mb.
səhifə42/117
tarix25.10.2017
ölçüsü8,63 Mb.
#13029
1   ...   38   39   40   41   42   43   44   45   ...   117

Other

  1. Decoder optimization


4.2.0.1.1.1.1.1.98(Discussed in joint meeting Video/Requirements.) m27613 was also presented (see Requirements report).

4.2.0.1.1.1.1.1.99m28114 Cross-segment decoding of HEVC bitstreams with SAD/PSNR information [Shunyao Li, Yao Lu, Gene Wen, Spencer Cheng]

The contribution described an alternative decoding process that differs from the ordinarly-specified decoding process for coded video content.

Assumption: Segments of video are encoded with time-varying bit rate / quality and served to the decoder client according to adaptive bit rate capacity.

If the last frame of previous segment is encoded by better quality, an intelligent decoder could use that in cases when the beginning of current segment is of lower quality.

Results were shown with graphics and videoconferencing (static) content, reporting about 0.3% PSNR improvement average.

Motion estimation at then decoder is necessary, which incurs an additional complexity burden on decoders.

In regard to the relation with / assumption of random access, IDR usage was assumed.

It was asked whether that decoding process would be normative. That was not proposed, but potential usage in the context of scalability is also mentioned. With complexity scalability, one could have drift.

It was asked why it is necessary to put PSNR data in the bitstream. This was done to determine the variation of quality. It was asked whether this could also be done by evaluating the QP values. If done that way, it was suggested that the whole approach could be done by non-normative means anyway.

In essence, this is non-normative processing after the normative decoding process, conceptually with some similarity with error concealment. It was not clear whether it is necessary to convey the PSNR parameters.

Further study was necessary before the benefit of sending additional data could be assessed.


  1. Wednesday Video plenary status review


The status was reviewed 14:30-16:30 on Wednesday.

Breakout report discussions included the following:



CDVS

  • Three different alternative solutions for keypoint detection (4 inputs)

  • Some people doubt whether the patent issue is resolved

  • This should be further investigated and if the issue is not resolved, hard facts should be brought in future inputs.

  • In the meantime, the technology should be further technically investigated and improved (in terms of performance and complexity, as usual)

  • Only one configuration in future (the previous “H” mode; the “S” mode is to be removed.

  • Various improvements: Global descriptor with less parameters (from 102 to 32), but compensated by using more clusters at high bit rates (and better tradeoff of bit rate assigned between global and local descriptor, such that overall performance is improved), no change at low bit rates

  • Software integration is challenging

  • The TM would still use SIFT, alternatives to be investigated in CE

  • This means that other elements of the chain will be investigated with SIFT, but it is already known that performance with alternatives would be similar

  • Precise description of the new keypoint detection methods has to be developed

  • Description of feature extraction is planned to be developed after the meeting

  • Delay by one more meeting cycle; review next meeting whether CDVS is ready for CD.

IVC:

  • 4 new tools added to ITM

  • ITM with new tools integrated will be compared vs. VP8 (CE1)

  • Comparison conditions are being defined

  • Another CE (CE2) for further improvements of ITM (with several sub-experiments)

  • Visual tests planned (informal this meeting, then either formal before or expert during next meeting)

Output doc planning was further confirmed.
  1. Closing plenary topics


A closing plenary of the Video SG was held Friday 9:00–12:15. Results are recorded in this section.
    1. Output docs


The following output documents were approved:

No.

Title

TBP

Available




14496-2 – Visual







13322

Text of ISO/IEC 14496-2:2004/DCOR5 (correction of rounding operation)

N

13/01/25




14496-4 – Conformance testing







13323

Request for ISO/IEC 14496-4:2004/Amd.41

N

13/01/25

13324

Text of ISO/IEC 14496-4:2004/PDAM41 Conformance testing of the MVC plus depth extension of AVC

N

13/02/01




14496-10 – Advanced Video Coding







13325

Text of ISO/IEC 14496-10:2012/DCOR1 (correction of MVC sub-bitstream extraction process)

N

13/01/25

13326

Disposition of Comments on ISO/IEC 14496-10:2012/DAM2

N

13/01/25

13327

Text of ISO/IEC 14496-10:2012/FDAM2 MVC extension for inclusion of depth maps

N

13/03/22

13328

Request for ISO/IEC 14496-10:2012/Amd.4

N

13/01/25

13329

Text of ISO/IEC 14496-10:2012/PDAM4 Additional Colour Space and Tone Mapping Descriptors

N

13/02/01

13330

Request for ISO/IEC 14496-10:2012/Amd.5

N

13/01/25

13331

Text of ISO/IEC 14496-10:2012/PDAM5 Multi-resolution Frame Compatible Stereo Coding

N

13/02/01

13332

Study Text of ISO/IEC 14496-10:2012/PDAM3 AVC compatible video-plus-depth extension

N

13/02/06

13342

3D-AVC Test Model 5

N

13/03/22




15938-13 – Compact Descriptors for Visual Search







13334

Text of Working Draft 3: Compact Descriptors for Visual Search

N

13/02/15

13335

Test Model 5: Compact Descriptors for Visual Search

N

13/02/15

13336

Core Experiments in Compact Descriptors for Visual Search

N

13/01/25




23002-4 – Media Tool Library







13337

WD1 of FU and FN descriptions for HEVC

N

13/01/25

13338

Core Experiment on Investigation of HEVC related FU and FN Implementations for RVC

N

13/01/25




23002-5 – Reconfigurable Media Coding Conformance and Reference Software







13339

Disposition of Comments on ISO/IEC DIS 23002-5

N

13/01/25

13340

Text of ISO/IEC FDIS 23002-5 Reconfigurable Media Coding Conformance and Reference Software

N

13/02/01




23008-2 – High Efficiency Video Coding







13341

Disposition of Comments on ISO/IEC DIS 23008-2

N

13/01/25

13333

Text of ISO/IEC FDIS 23008-2 High Efficiency Video Coding

N

13/03/01

13343

High Efficiency Video Coding (HEVC) Test Model 10 (HM 10) Encoder Description

N

13/04/01

13344

Request for ISO/IEC 23008-2:201x/Amd.1

N

13/01/25

13345

Text of ISO/IEC 23008-2:201x/PDAM1 HEVC Range Extensions

N

13/03/22

13346

Request for ISO/IEC 23008-2:201x/Amd.2

N

13/01/25

13347

Text of ISO/IEC 23008-2:201x/PDAM2 HEVC Multiview Extensions

N

13/03/22

13348

3D-HEVC Test Model 3

N

13/03/22

13349

Scalable HEVC (SHVC) Working Draft 1

N

13/03/22

13350

Scalable HEVC (SHVC) Test Model 1 (SHM 1)

N

13/02/15




23008-5 – Conformance and Reference Software for High Efficiency Video Coding







13351

Request for Subdivision: ISO/IEC 23008-5

N

13/01/25

13352

Text of ISO/IEC CD 23008-5 HEVC Conformance Testing and Reference Software

N

13/03/22




Exploration – Internet Video Coding







13353

Internet Video Coding Test Model (ITM) v 4.0

N

13/02/20

13354

IVC Core Experiment CE1: Overall Codec Testing

N

13/01/25

13355

IVC Core Experiment CE2: Improvements of ITM

N

13/01/25




Liaison Statements







13356

Liaison Statement template to various organizations re HEVC development

N

13/01/25

13357

Liaison Statement template to various organizations re MFC development

N

13/01/25

13358

Liaison Statement to EBU re test sequences

N

13/01/25




National Bodies







13359

Responses to NBs re HEVC

N

13/01/25




    1. Yüklə 8,63 Mb.

      Dostları ilə paylaş:
1   ...   38   39   40   41   42   43   44   45   ...   117




Verilənlər bazası müəlliflik hüququ ilə müdafiə olunur ©muhaz.org 2024
rəhbərliyinə müraciət

gir | qeydiyyatdan keç
    Ana səhifə


yükləyin