Research map / Architectures and efficiency / Attention and transformers
OAPT: Offset-Aware Partition Transformer for Double JPEG Artifacts Removal
Qiao Mo, Yukang Ding, Jinhua Hao, Qiang Zhu, Ming Sun, Chao Zhou, Feiyu Chen, Shuyuan Zhu
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 pattern-clustering attention with offset predictor to SwinIR backbone | SwinIR · Classic5/LIVE1/BSDS500 grayscale double JPEG | PSNR (dB): about +0.06 dB average over SwinIR; 1.5M extra parameters | Improved |
| use ground-truth offsets instead of predicted offsets (ablation) | OAPT · LIVE1 (70,40,5,1) | PSNR (dB): 29.48 -> 29.53 | Improved |
| plug pattern clustering into OCAB of HAT-S with ground-truth offsets | HAT-S · double JPEG restoration | PSNR (dB): +0.03 dB average | Improved |
| parallel hybrid attention instead of serial hybrid attention (ablation) | OAPT HPAB · LIVE1 (30,50,4,4) | PSNR (dB) | Hurt |
| use uniform sparse attention only instead of serial hybrid (dense plus pattern-clustered) (ablation) | OAPT HPAB · LIVE1 (30,50,4,4) | PSNR (dB) | Hurt |
| remove pattern clustering (uniform dense attention equals RSTB) (ablation) | OAPT HPAB · LIVE1 (30,50,4,4) | PSNR (dB) | Hurt |
| offset prediction when QF1 >= QF2 | OAPT offset predictor · LIVE1 (90,10,2,0) | offset accuracy: accuracy 97% at (10,90,0,2) vs 0% at (90,10,2,0) | Hurt |
| train on double compression only then test on single compression (QF 10-50) | OAPT · LIVE1 single compression | PSNR (dB): 29.75 -> 29.77 (QF=10 versus SwinIR) | Improved |
| use ART-style uniform sparse attention instead of OAPT | ART · Classic5 (50,30,4,4) | PSNR (dB): OAPT 32.97 vs ART 33.02 | Mixed |
Closest papers by meaning
- Towards Flexible Blind JPEG Artifacts Removal (ICCV 2021 · similarity 0.86)
- Compression-Aware One-Step Diffusion Model for JPEG Artifact Removal (ICCV 2025 · similarity 0.86)
- SODiff:Semantic-Oriented Diffusion Model for JPEG Compression Artifacts Removal (AAAI 2026 · similarity 0.86)
- DCTMamba: Advancing JPEG Image Restoration Through Long-Sequence Modeling and Adaptive Frequency Strategy (AAAI 2025 · similarity 0.83)
- CosAE: Learnable Fourier Series for Image Restoration (NeurIPS 2024 · similarity 0.82)
- JDEC: JPEG Decoding via Enhanced Continuous Cosine Coefficients (CVPR 2024 · similarity 0.82)
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.
