Research map / Image generation / Image restoration
Unpaired Deep Image Deraining Using Dual Contrastive Learning
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 DCD-GAN instead of the strongest listed baseline | unpaired image deraining · Test100 | PSNR: 24.37 -> 25.61 | Helped |
| Use DCD-GAN instead of the strongest listed baseline | unpaired image deraining · Test1200 | SSIM: 0.934 -> 0.937 | Helped |
| Use DCD-GAN instead of the strongest listed baseline | unpaired image deraining · Test1400 | PSNR: 32.03 -> 31.65 | Hurt |
| Use DCD-GAN instead of the strongest listed baseline | unpaired image deraining · Test1000 | SSIM: 0.950 -> 0.943 | Hurt |
| Use DCD-GAN instead of the strongest listed baseline | real-world image deraining · Real-world rainy images | NIQE: 4.263 -> 4.032 | Helped |
| Add adversarial loss (ablation) | DCD-GAN · Test100 | PSNR: 24.03 -> 25.61 | Helped |
| Add color cycle-consistency loss (ablation) | DCD-GAN · Test100 | PSNR: 25.08 -> 25.61 | Helped |
| Use color cycle-consistency loss instead of standard cycle-consistency loss (ablation) | DCD-GAN · Test100 | PSNR: 25.32 -> 25.61 | Helped |
| Add contrastive loss (ablation) | DCD-GAN · Test100 | PSNR: 22.84 -> 25.61 | Helped |
| Add frequency loss (ablation) | DCD-GAN · Test100 | PSNR: 25.40 -> 25.61 | Helped |
| Use internal and external negatives instead of internal negatives only (ablation) | DCD-GAN · Test100 | PSNR: 25.47 -> 25.61 | Helped |
| Use internal and external negatives instead of internal negatives only (ablation) | DCD-GAN · Test100 | SSIM: 0.815 -> 0.813 | Hurt |
Closest papers by meaning
- Unsupervised Deraining: Where Contrastive Learning Meets Self-Similarity (CVPR 2022 · similarity 0.88)
- Channel Consistency Prior and Self-Reconstruction Strategy Based Unsupervised Image Deraining (CVPR 2025 · similarity 0.87)
- End-to-End Unpaired Image Denoising with Conditional Adversarial Networks (AAAI 2020 · similarity 0.86)
- Rethinking Multi-Scale Representations in Deep Deraining Transformer (AAAI 2024 · similarity 0.85)
- Toward Improving Robustness and Accuracy in Unsupervised Domain Adaptation (AAAI 2025 · similarity 0.83)
- Learning Image Demoiréing from Unpaired Real Data (AAAI 2024 · similarity 0.83)
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.