Video coding standards k. R. Rao, Do Nyeon Kim J. J. Hwang Springer 2014

SUBJECTIVE EVALUATON OF COMPRESSION ALGORITHMS AND STANDARDS

SE.1 P. Hanhart and T. Ebrahimi, “Calculation of average coding efficiency based on subjective quality scores”, J. VCIR, vol. 25, pp.555-564, April 2014. This is a very interesting and valuable paper on subjective quality and testing. The references listed at the end are highly useful. A MATLAB implementation of the proposed model can be downloaded from http://mmspg.epfl.ch/scenic This paper can lead to several projects (EE5359 Multimedia Processing).

SE.2 H.R. Wu et al, “Perceptual visual signal compression and transmission”, Proc. IEEE, vol. 101, pp.2025-2043, Sept. 2013.

SE.3 C. Deng et al (Editors), “Visual Signal Quality Assessment: Quality of Experience (QoE)”, Springer, 2015.

SE.4 T.K. Tan et al, “Video Quality Evaluation Methodology and Verification Testing of HEVC Compression Performance”, IEEE Trans. CSVT, vol. 26, pp.76-90, Jan. 2016. Abstract of this paper is reproduced here for ready reference.

Abstract— The High Efficiency Video Coding (HEVC) standard (ITU-T H.265 and ISO/IEC 23008-2) has been developed with the main goal of providing significantly improved video compression compared with its predecessors. In order to evaluate this goal, verification tests were conducted by the Joint Collaborative Team on Video Coding of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29. This paper presents the subjective and objective results of a verification test in which the performance of the new standard is compared with its highly successful predecessor, the Advanced Video Coding (AVC) video compression standard (ITU-T H.264 and ISO/IEC 14496-10). The test used video sequences with resolutions ranging from 480p up to ultra-high definition, encoded at various quality levels using the HEVC Main profile and the AVC High profile. In order to provide a clear evaluation, this paper also discusses various aspects for the analysis of the test results. The tests showed that bit rate savings of 59% on average can be achieved by HEVC for the same perceived video quality, which is higher than a bit rate saving of 44% demonstrated with the PSNR objective quality metric. However, it has been shown that the bit rates required to achieve good quality of compressed content, as well as the bit rate savings relative to AVC, are highly dependent on the characteristics of the tested content.

This paper has many valuable references including subjective quality assessment methods recommended by ITU-T.

SE5 J.-S. Lee and T. Ebrahimi, “Perceptual video compression: A survey,” IEEE J. Selected Topics on Signal Process., vol. 6, no. 6, pp. 684–697, Oct. 2012.
SE6 P. Hanhart et al, “Subjective quality evaluation of the upcoming HEVC video compression standard”, SPIE Applications of digital image processing XXXV, vol. 8499, paper 8499-30, Aug. 2012.

SE7 W. Lim and J.C. Kuo, “Perceptual video quality metric: a survey”, J. VCIR, vol.22, pp.297-312, 2011.

SE8 F. Zhang and D.R. Bull, “A Perception-based Hybrid Model for Video Quality Assessment “, IEEE Trans. CSVT, vol.26, pp.1017-1028, June 2016.

SE9 Y. Li et al, "No-reference image quality assessment using statistical characterization in the shearlet domain." Signal Processing: Image Communication, vol. 29, pp.748-759, July 2014.

SE10 Y. Li et al, "No reference image quality assessment with shearlet transform and deep neural networks," Neurocomputing, vol.154, pp. 94-109, 2015.

SE11 Y. Li et al, "No Reference Video Quality Assessment with 3D shearlet Transform and Convolutional Neural Networks," IEEE Trans. on CSVT, vol. 26, pp.1044-1047, June 2016.

SE12 Y. Li, "No-reference image quality assessment using shearlet transform and stacked auto encoders," IEEE ISCAS, pp. 1594-1597, May 2015.

SE13 T. Daede, “Video codec testing and quality measurement," https://tools.ietf.org/html/ draft-daede-netvc-testing, 2015.

SE14 ITU-R BT2022 (2012) General viewing conditions for subjective assessment of quality of SDTV and HDTV television pictures on at panel displays. International Telecommunication Union.

SE15 ITU-R BT500-13 (2012) Methodology for the subjective assessment of the quality of television pictures. International Telecommunication Union.

SE16 ITU-R BT500-13 (2012) Methodology for the subjective assessment of the quality of television pictures. International Telecommunication Union.

SE17 W. Lin and C.C. Jay-Kuo, “Perceptual visual quality metrics: A survey”, J. Visual Communication Image Representation, vol.22, pp. 297-312, May 2011.

SE18. C.-C. Jay Kuo, “Perceptual coding: Hype or Hope?”, 8^th International Conference on QoMEX, Keynote, Lisbon, Portugal, 6-8 June 2016.

SE19. S.-H. Bae and M. Kim, “DCT-QM; A DCT-Based Quality Degradation Metric for Image Quality Optimization Problems ”, IEEE Trans. IP, vol.25, pp. 4916-4930, Oct. 2016. This has several references related to image quality assessment.

SE20. IEEE DATA COMPRESSION CONFERENCE (DCC) Snowbird, Utah, April 4 - 7, 2017. http://www.cs.brandeis.edu/~dcc, Keynote address

"Video Quality Metrics" by Scott Daly, Senior Member Technical Staff, Dolby Laboratories.

SE21 F.M. Moss et al, “On the Optimal Presentation Duration for Subjective Video Quality Assessment”, IEEE Trans. CSVT, vol.26, pp.1977-1987, Nov. 2016.
Apart from the comprehensive list of references related to subjective VQA, based on the DSCQS

tests on video sequences the authors [SE21] conclude

“There is a small but significant increase in accuracy if sequences are increased from 1.5 to 5, 7, or 10 s.

2) This effect becomes stronger if the difference in distortion between the reference and test video is reduced.

3) The main effect remains consistent between different but temporally consistent source videos.

4) Observers feel just as confident assessing the quality of videos that are 5 s as ones that are 10 s.

The practical implications of these findings are significant.

Our results indicate that critical observations of video quality do not significantly change if 10-s sequences are exchanged for 7-, or indeed, 5-s sequences.” Limitations on these findings are elaborated in the conclusions.
SE-P1 Bae and Kim have developed the DCT-Q Based Quality Degradation Metric for Image Quality Optimization Problem that has both high consistency with perceived quality and mathematically desirable properties. See [SE19] for details. The authors conclude that the DCT-QM can effectively be used not only for objective IQA tasks with unknown distortion types but also for image quality optimization problems with known distortion types. They also state “We plan to apply our DCT-QM for image/video coding and processing applications for the improvement of coding efficiency and perceptual visual quality.” Explore this. Note that DCT-QM serves for still images only.

SE-P2 See SE-P1. Note that the performance comparison of HEVC (DCT-QM based HM16.0and conventional HM16.0 for low delay configuration) is shown in Fig.9 for test sequences BasketballDrill, BQ Mall, PartyScene and RaceHorse). See also Figs. 10 and 11. Please extend this comparison to HM 16.0 random access configuration.

SE-P3 See [SE19] The authors state “Full application of DCT-QM for specific codecs require extending DCT-QM for specific transform block sizes, which will be part of our future work”. They imply other than 4x4-sized DCT kernel i.e., 8x8, 16x16, 32x32 and even 64x64. Consider specifically how the different sized DCT kernels influence the DCT-QM. Consider also the implementation complexity.