Research map / Image generation / Image restoration
JPEG Artifacts Removal via Contrastive Representation Learning
Xi Wang, Xueyang Fu, Yurui Zhu, Zheng-Jun Zha
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add the contrastively pre-trained compression quality encoder (ablation) | blind JPEG artifacts removal network · Classic5 Q10 | PSNR (dB): 29.97 -> 30.16 | Improved |
| add pre-training of the compression quality encoder (vs joint training without it) (ablation) | blind JPEG artifacts removal network · Classic5 Q10 | PSNR (dB): 30.04 -> 30.16 | Improved |
| add invertible neural module fusion instead of concat and conv (ablation) | blind JPEG artifacts removal network · Classic5 Q10 | PSNR (dB): 30.11 -> 30.16 | Improved |
| add INM and EFFM fusion instead of concat and conv (ablation) | blind JPEG artifacts removal network · Classic5 Q10 | PSNR (dB): 30.09 -> 30.16 | Improved |
| add FFT loss (ablation) | blind JPEG artifacts removal network · Classic5 Q10 | PSNR (dB): 30.03 -> 30.16 | Improved |
| add FFT loss (ablation) | blind JPEG artifacts removal network · Classic5 Q40 | PSNR (dB): 34.32 -> 34.43 | Improved |
| add compression quality encoder (ablation) | blind JPEG artifacts removal network · Classic5 Q40 | PSNR (dB): 34.31 -> 34.43 | Improved |
| use single contrastive model versus FBCNN | blind JPEG artifacts removal network · LIVE1 Q10 | PSNR (dB): 29.75 -> 29.80 | Improved |
| use contrastive model versus FBCNN on unseen quality | blind JPEG artifacts removal network · Classic5 Q15 | PSNR (dB): 31.42 -> 31.37 | No clear effect |
| use contrastive model versus FBCNN on color ICB | blind JPEG artifacts removal network · ICB color Q10 | PSNR (dB): 32.18 -> 32.05 | Hurt |
Closest papers by meaning
- Compression-Aware One-Step Diffusion Model for JPEG Artifact Removal (ICCV 2025 · similarity 0.92)
- Towards Flexible Blind JPEG Artifacts Removal (ICCV 2021 · similarity 0.91)
- Learning Dual Priors for JPEG Compression Artifacts Removal (ICCV 2021 · similarity 0.91)
- JPEG Artifacts Reduction via Deep Convolutional Sparse Coding (ICCV 2019 · similarity 0.90)
- DDR: Exploiting Deep Degradation Response as Flexible Image Descriptor (NeurIPS 2024 · similarity 0.84)
- Frequency-Biased Synergistic Design for Image Compression and Compensation (CVPR 2025 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
