DEEP PERCEPTUAL IMAGE QUALITY ASSESSMENT FOR COMPRESSION

Khaydaraliyeva K. . & Ergashova D. .


Annotatsiya

Lossy Image compression is necessary for efficient storage and transfer of data. Typically the trade-off between bit-rate and quality determines the optimal compression level. This makes the image quality metric an integral part of any imag- ing system. While the existing full-reference metrics such as PSNR and SSIM may be less sensitive to perceptual quality, the recently introduced learning methods may fail to general- ize to unseen data. In this paper we propose the largest image compression quality dataset to date with human perceptual preferences, enabling the use of deep learning, and we de- velop a full reference perceptual quality assessment metric for lossy image compression that outperforms the existing state-of-the-art methods. We show that the proposed model can effectively learn from thousands of examples available in the new dataset, and consequently it generalizes better to other unseen datasets of human perceptual preference. The CIQA dataset can be found at https://github.com/google- research/google-research/tree/master/CIQA


Adabiyotlar

  1. Muzhir Al-Ani and Fouad Awad, “The jpeg image com- pression algorithm,” Int. Journal of Advances in En- gineering And Technology, vol. 6, pp. 1055–1062, 05 2013.
  2. Xiyang Luo, H. Talebi, Feng Yang, Michael Elad, and P. Milanfar, “The rate-distortion-accuracy tradeoff: Jpeg case study,” DCC, 2020.
  3. Nikolay Ponomarenko, Lina Jin, Oleg Ieremeiev, Vladimir Lukin, Karen Egiazarian, Jaakko Astola, Benoit Vozel, Kacem Chehdi, Marco Carli, Federica Battisti, et al., “Image database tid2013: Peculiarities, results and perspectives,” Sig. Proces.: Image Comm., vol. 30, pp. 57–77, 2015.
  4. Zhou Wang, Alan C Bovik, Hamid R Sheikh, and Eero P Simoncelli, “Image quality assessment: from error visi- bility to structural similarity,” IEEE TIP, vol. 13, no. 4, pp. 600–612, 2004.
  5. Zhou Wang, Eero P Simoncelli, and Alan C Bovik, “Multiscale structural similarity for image quality as- sessment,” in ACSSC. IEEE, 2003, vol. 2, pp. 1398– 1402.
  6. Jim Nilsson and Tomas Akenine-Mo¨ller, “Understand- ing ssim,” arXiv preprint arXiv:2006.13846, 2020.
  7. Richard Zhang, Phillip Isola, Alexei A Efros, Eli Shechtman, and Oliver Wang, “The unreasonable ef- fectiveness of deep features as a perceptual metric,” in CVPR, 2018, pp. 586–595.
  8. Hossein Talebi and Peyman Milanfar, “Nima: Neural image assessment,” IEEE TIP, vol. 27, no. 8, pp. 3998– 4011, 2018.
  9. “Usage statistics of jpeg for websites,” https://w3techs.com/technologies/details/im-jpeg, Accessed: 2021-01-25.
  10. L. Cormack H.R. Sheikh, Z.Wang and A.C. Bovik, “Live image quality assessment database release 2,” 2006, Accessed: 2020-07-25.
  11. X. Zhang, W. Lin, S. Wang, J. Liu, S. Ma, and W. Gao, “Fine-grained quality assessment for compressed im- ages,” IEEE TIP, vol. 28, no. 3, pp. 1163–1175, 2019.
  12. N Krishnaraj, Mohamed Elhoseny, M Thenmozhi, Mah- moud M Selim, and K Shankar, “Deep learning model for real-time image compression in internet of underwa- ter things (iout),” J. of Real-Time Image Proc., vol. 17, no. 6, pp. 2097–2111, 2020.
  13. J. Uthayakumar, M. Elhoseny, and K. Shankar, “Highly reliable and low-complexity image compression scheme using neighborhood correlation sequence algorithm in wsn,” IEEE Trans. on Reliability, vol. 69, no. 4, pp. 1398–1423, 2020.
  14. Siu-Wai Wu and Allen Gersho, “Rate-constrained picture-adaptive quantization for jpeg baseline coders,” in ICASSP. IEEE, 1993, vol. 5, pp. 389–392.
  15. Ties van Rozendaal, Guillaume Sautiere, and Taco S. Cohen, “Lossy compression with distortion constrained optimization,” in CVPR, June 2020.
  16. Li-Heng Chen, Christos G Bampis, Zhi Li, Andrey Norkin, and Alan C Bovik, “Perceptually opti- mizing deep image compression,” arXiv preprint arXiv:2007.02711, 2020.
  17. Naila Murray, Luca Marchesotti, and Florent Perronnin, “Ava: A large-scale database for aesthetic visual analy- sis,” in CVPR. IEEE, 2012, pp. 2408–2415.
  18. Vlad Hosu, Hanhe Lin, Tamas Sziranyi, and Dietmar Saupe, “Koniq-10k: An ecologically valid database for deep learning of blind image quality assessment,” IEEE TIP, vol. 29, pp. 4041–4056, 2020.
  19. Chen Kang, Giuseppe Valenzise, and Fre´de´ric Du- faux, “Predicting subjectivity in image aesthetics as- sessment,” in MMSP. IEEE, 2019, pp. 1–6.
  20. Hossein Talebi, Ehsan Amid, Peyman Milanfar, and Manfred K Warmuth, “Rank-smoothed pairwise learn- ing in perceptual quality assessment,” in ICIP. IEEE, 2020, pp. 3413–3417.
  21. Rafal Mantiuk, Anna Tomaszewska, and Radoslaw Mantiuk, “Comparison of four subjective methods for image quality assessment,” Comp. Graph. Forum, vol. 31, 11 2012.
  22. Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei, “Imagenet: A large-scale hierarchical image database,” in CVPR. IEEE, 2009, pp. 248–255.
  23. Mingxing Tan and Quoc Le, “Efficientnet: Rethinking model scaling for convolutional neural networks,” in ICML. PMLR, 2019, pp. 6105–6114.
  24. Gao Huang, Zhuang Liu, Laurens Van Der Maaten, and Kilian Q Weinberger, “Densely connected convolutional networks,” in CVPR, 2017, pp. 4700–4708.
  25. Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun, “Identity mappings in deep residual networks,” in ECCV. Springer, 2016, pp. 630–645.