Research map / Label-efficient and robust learning / Domain adaptation and generalization
Physically Disentangled Intra- and Inter-Domain Adaptation for Varicolored Haze Removal
Yi Li, Yi Chang, Yan Gao, Changfeng Yu, Luxin Yan
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 |
|---|---|---|---|
| Use PDI2A instead of the strongest listed baseline | Real varicolored haze removal · Real varicolored haze | NIQE: 5.83 -> 5.61 | Improved |
| Use PDI2A instead of the strongest listed baseline | Real varicolored haze removal · Real varicolored haze | User Study: 5.73 -> 6.80 | Improved |
| Use PDI2A instead of Semi-Dehazing | Synthetic varicolored haze removal · Synthetic varicolored haze | PSNR: 28.70 -> 30.13 | Improved |
| Use PDI2A instead of Semi-Dehazing | Synthetic varicolored haze removal · Synthetic varicolored haze | SSIM: 0.9638 -> 0.9670 | Improved |
| Add intra-domain adaptation for color correction (ablation) | PDI2A · Synthetic varicolored haze | PSNR: 29.20 -> 30.13 | Improved |
| Add inter-domain adaptation for synthetic-real translation (ablation) | PDI2A · Synthetic varicolored haze | PSNR: 29.07 -> 30.13 | Improved |
| Add disentangle consistency loss (ablation) | PDI2A · Synthetic varicolored haze | PSNR: 28.47 -> 30.13 | Improved |
| Add reconstruction consistency loss (ablation) | PDI2A · Synthetic varicolored haze | PSNR: 29.94 -> 30.13 | Improved |
| Train intra- and inter-domain adaptation jointly rather than separately (ablation) | PDI2A · Synthetic varicolored haze | PSNR: 27.95 -> 30.13 | Improved |
| Pretrain on synthetic data before joint finetuning (ablation) | PDI2A · Synthetic varicolored haze | PSNR: 24.95 -> 30.13 | Improved |
Closest papers by meaning
- PRISM: Rethinking Atmospheric Scattering Reconstruction as a Unified Understanding and Restoration Model for Real-world Dehazing (NeurIPS 2026 · similarity 0.91)
- Disentanglement-wise Image Dehazing through Cross-Domain Manifold Consensus (CVPR 2026 · similarity 0.91)
- Learning Hazing to Dehazing: Towards Realistic Haze Generation for Real-World Image Dehazing (CVPR 2025 · similarity 0.88)
- Guided Real Image Dehazing Using YCbCr Color Space (AAAI 2025 · similarity 0.87)
- Tokenize Image Patches: Global Context Fusion for Effective Haze Removal in Large Images (CVPR 2025 · similarity 0.87)
- Depth-Centric Dehazing and Depth-Estimation from Real-World Hazy Driving Video (AAAI 2025 · similarity 0.86)
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.
