Papers
Topics
Authors
Recent
Assistant
AI Research Assistant
Well-researched responses based on relevant abstracts and paper content.
Custom Instructions Pro
Preferences or requirements that you'd like Emergent Mind to consider when generating responses.
Gemini 2.5 Flash
Gemini 2.5 Flash 83 tok/s
Gemini 2.5 Pro 54 tok/s Pro
GPT-5 Medium 21 tok/s Pro
GPT-5 High 20 tok/s Pro
GPT-4o 103 tok/s Pro
Kimi K2 205 tok/s Pro
GPT OSS 120B 456 tok/s Pro
Claude Sonnet 4 35 tok/s Pro
2000 character limit reached

Enhancing Quality of Compressed Images by Mitigating Enhancement Bias Towards Compression Domain (2402.17200v3)

Published 27 Feb 2024 in cs.CV and eess.IV

Abstract: Existing quality enhancement methods for compressed images focus on aligning the enhancement domain with the raw domain to yield realistic images. However, these methods exhibit a pervasive enhancement bias towards the compression domain, inadvertently regarding it as more realistic than the raw domain. This bias makes enhanced images closely resemble their compressed counterparts, thus degrading their perceptual quality. In this paper, we propose a simple yet effective method to mitigate this bias and enhance the quality of compressed images. Our method employs a conditional discriminator with the compressed image as a key condition, and then incorporates a domain-divergence regularization to actively distance the enhancement domain from the compression domain. Through this dual strategy, our method enables the discrimination against the compression domain, and brings the enhancement domain closer to the raw domain. Comprehensive quality evaluations confirm the superiority of our method over other state-of-the-art methods without incurring inference overheads.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (52)
  1. NTIRE 2017 challenge on single image super-resolution: Dataset and study. In 2017 IEEE conference on computer vision and pattern recognition workshops, CVPR workshops 2017, honolulu, HI, USA, july 21-26, 2017, pages 1122–1131. IEEE Computer Society, 2017. tex.bibsource: dblp computer science bibliography, https://dblp.org tex.biburl: https://dblp.org/rec/conf/cvpr/AgustssonT17.bib tex.timestamp: Fri, 09 Apr 2021 18:48:31 +0200.
  2. Towards Principled Methods for Training Generative Adversarial Networks. In International Conference on Learning Representations, 2017.
  3. Layer Normalization, 2016. arXiv:1607.06450 [cs, stat].
  4. Fabrice Bellard. Better portable graphics (BPG), 2018.
  5. Gisle Bjontegaard. Calculation of average PSNR differences between RD-curves. VCEG-M33, 2001.
  6. The perception-distortion tradeoff. In 2018 IEEE/CVF conference on computer vision and pattern recognition. IEEE, 2018.
  7. The 2018 PIRM Challenge on Perceptual Image Super-Resolution. In Proceedings of the European Conference on Computer Vision (ECCV) Workshops, 2018.
  8. TOPIQ: A Top-down Approach from Semantics to Distortions for Image Quality Assessment, 2023. arXiv:2308.03060 [cs].
  9. Image Quality Assessment: Unifying Structure and Texture Similarity. IEEE Transactions on Pattern Analysis and Machine Intelligence, pages 1–1, 2020. arXiv:2004.07728 [cs].
  10. Compression artifacts reduction by a deep convolutional network. In 2015 IEEE international conference on computer vision (ICCV). IEEE, 2015.
  11. IEGAN: Multi-Purpose Perceptual Quality Image Enhancement Using Generative Adversarial Network. In 2019 IEEE Winter Conference on Applications of Computer Vision (WACV), pages 11–20, 2019. ISSN: 1550-5790.
  12. Generative adversarial nets. In Advances in neural information processing systems. Curran Associates, Inc., 2014.
  13. MFQE 2.0: A New Approach for Multi-frame Quality Enhancement on Compressed Video. IEEE Transactions on Pattern Analysis and Machine Intelligence, 43(3):949–963, 2021. arXiv:1902.09707 [cs].
  14. Building dual-domain representations for compression artifacts reduction. In Computer vision – ECCV 2016, pages 628–644. Springer International Publishing, 2016.
  15. Toward convolutional blind denoising of real photographs. In 2019 IEEE/CVF conference on computer vision and pattern recognition (CVPR). IEEE, 2019.
  16. Deep residual learning for image recognition. In 2016 IEEE conference on computer vision and pattern recognition (CVPR). IEEE, 2016.
  17. GANs trained by a two time-scale update rule converge to a local nash equilibrium. In Proceedings of the 31st international conference on neural information processing systems, pages 6629–6640, Red Hook, NY, USA, 2017. Curran Associates Inc. Number of pages: 12 Place: Long Beach, California, USA.
  18. Domo Inc. Data Never Sleeps 8.0: How much data is generated every minute?, 2020.
  19. International Telecommunication Union Communication Standardization Sector (ITU-T). P.10 : Vocabulary for performance, quality of service and quality of experience, 2017.
  20. Alexia Jolicoeur-Martineau. The relativistic discriminator: a key element missing from standard GAN. In 7th International Conference on Learning Representations, ICLR 2019, New Orleans, LA, USA, May 6-9, 2019. OpenReview.net, 2019.
  21. MUSIQ: Multi-Scale Image Quality Transformer. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), pages 5148–5157, 2021.
  22. Towards the Perceptual Quality Enhancement of Low Bit-rate Compressed Images. In 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), pages 565–569, Los Alamitos, CA, USA, 2020. IEEE Computer Society.
  23. Attentions Help CNNs See Better: Attention-Based Hybrid Image Quality Assessment Network. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, pages 1140–1149, 2022.
  24. Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network. In 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 105–114, Los Alamitos, CA, USA, 2017. IEEE Computer Society. ISSN: 1063-6919.
  25. IID-GAN: an IID Sampling Perspective for Regularizing Mode Collapse. In Proceedings of the Thirty-Second International Joint Conference on Artificial Intelligence, pages 3929–3938, Macau, SAR China, 2023. International Joint Conferences on Artificial Intelligence Organization.
  26. An overview of JPEG-2000. In Proceedings DCC 2000. Data compression conference. IEEE Comput. Soc, 2000.
  27. Conditional Generative Adversarial Nets, 2014. arXiv:1411.1784 [cs, stat].
  28. Making a “Completely Blind” Image Quality Analyzer. IEEE Signal Processing Letters, 20(3):209–212, 2013. Conference Name: IEEE Signal Processing Letters.
  29. Spectral Normalization for Generative Adversarial Networks. In International Conference on Learning Representations, 2018.
  30. A U-Net Based Discriminator for Generative Adversarial Networks. In 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 8204–8213, Los Alamitos, CA, USA, 2020. IEEE Computer Society.
  31. Study of subjective and objective quality assessment of video. IEEE Transactions on Image Processing, 19(6):1427–1441, 2010. Publisher: Institute of Electrical and Electronics Engineers (IEEE).
  32. Review of Postprocessing Techniques for Compression Artifact Removal. Journal of Visual Communication and Image Representation, 9(1):2–14, 1998.
  33. Very deep convolutional networks for large-scale image recognition. In 3rd international conference on learning representations, ICLR 2015, san diego, CA, USA, may 7-9, 2015, conference track proceedings, 2015. tex.bibsource: dblp computer science bibliography, https://dblp.org tex.biburl: https://dblp.org/rec/journals/corr/SimonyanZ14a.bib tex.timestamp: Wed, 17 Jul 2019 10:40:54 +0200.
  34. Blindly Assess Image Quality in the Wild Guided by a Self-Adaptive Hyper Network. In 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 3664–3673, 2020. ISSN: 2575-7075.
  35. Overview of the high efficiency video coding (HEVC) standard. IEEE Transactions on Circuits and Systems for Video Technology, 22(12):1649–1668, 2012. Publisher: Institute of Electrical and Electronics Engineers (IEEE).
  36. Graphs: Theory and Algorithms. Wiley, 1992.
  37. NTIRE 2017 challenge on single image super-resolution: Methods and results. In 2017 IEEE conference on computer vision and pattern recognition workshops, CVPR workshops 2017, honolulu, HI, USA, july 21-26, 2017, pages 1110–1121. IEEE Computer Society, 2017. tex.bibsource: dblp computer science bibliography, https://dblp.org tex.biburl: https://dblp.org/rec/conf/cvpr/TimofteAG0ZLSKN17.bib tex.timestamp: Thu, 21 Apr 2022 09:15:18 +0200.
  38. G.K. Wallace. The JPEG still picture compression standard. IEEE Transactions on Consumer Electronics, 38(1):xviii–xxxiv, 1992. Publisher: Institute of Electrical and Electronics Engineers (IEEE).
  39. Exploring CLIP for Assessing the Look and Feel of Images. Proceedings of the AAAI Conference on Artificial Intelligence, 37(2):2555–2563, 2023. Number: 2.
  40. A novel deep learning-based method of improving coding efficiency from the decoder-end for HEVC. In 2017 data compression conference (DCC). IEEE, 2017.
  41. ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks. In Proceedings of the European Conference on Computer Vision (ECCV) Workshops, 2018.
  42. Real-ESRGAN: Training Real-World Blind Super-Resolution With Pure Synthetic Data. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Workshops, pages 1905–1914, 2021.
  43. D3: Deep dual-domain based fast restoration of JPEG-Compressed images. In 2016 IEEE conference on computer vision and pattern recognition (CVPR). IEEE, 2016.
  44. Wikipedia contributors. Triangle inequality — Wikipedia, The Free Encyclopedia, 2023.
  45. Early exit or not: Resource-efficient blind quality enhancement for compressed images. In Computer vision - ECCV 2020 - 16th european conference, glasgow, UK, august 23-28, 2020, proceedings, part XVI, pages 275–292. Springer, 2020. tex.bibsource: dblp computer science bibliography, https://dblp.org tex.biburl: https://dblp.org/rec/conf/eccv/XingXLG20.bib tex.timestamp: Thu, 17 Feb 2022 16:43:16 +0100.
  46. DAQE: Enhancing the Quality of Compressed Images by Exploiting the Inherent Characteristic of Defocus. IEEE Transactions on Pattern Analysis and Machine Intelligence, pages 1–17, 2023. Conference Name: IEEE Transactions on Pattern Analysis and Machine Intelligence.
  47. Multi-frame quality enhancement for compressed video. In 2018 IEEE/CVF conference on computer vision and pattern recognition. IEEE, 2018.
  48. Multi-stage progressive image restoration. In IEEE conference on computer vision and pattern recognition, CVPR 2021, virtual, june 19-25, 2021, pages 14821–14831. Computer Vision Foundation / IEEE, 2021. tex.bibsource: dblp computer science bibliography, https://dblp.org tex.biburl: https://dblp.org/rec/conf/cvpr/ZamirA0HK0021.bib tex.timestamp: Mon, 30 Aug 2021 17:00:27 +0200.
  49. Learning Deep CNN Denoiser Prior for Image Restoration. In 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 2808–2817, Los Alamitos, CA, USA, 2017. IEEE Computer Society. ISSN: 1063-6919.
  50. The unreasonable effectiveness of deep features as a perceptual metric. In 2018 IEEE/CVF conference on computer vision and pattern recognition. IEEE, 2018a.
  51. Residual Dense Network for Image Super-Resolution. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018b.
  52. Progressive Training of A Two-Stage Framework for Video Restoration. In 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), pages 1023–1030, New Orleans, LA, USA, 2022. IEEE.
Citations (2)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now
Lightbulb On Streamline Icon: https://streamlinehq.com

Continue Learning

We haven't generated follow-up questions for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Generate Now
List To Do Tasks Checklist Streamline Icon: https://streamlinehq.com

Collections

Sign up for free to add this paper to one or more collections.

User Circle Single Streamline Icon: https://streamlinehq.com Sign Up