Research map / Image generation / Image restoration
Contrastive Learning for Compact Single Image Dehazing
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 |
|---|---|---|---|
| add adaptive mixup to the base network (ablation) | AECR-Net · RESIDE | PSNR: 33.85 -> 34.04 | Helped |
| add adaptive mixup to the base network with DFE (ablation) | AECR-Net · RESIDE | PSNR: 35.50 -> 36.20 | Helped |
| replace skip connection with adaptive mixup in the DFE network (ablation) | AECR-Net · RESIDE | PSNR: 35.59 -> 36.20; 0.61dB | Helped |
| add DFE to the base network (ablation) | AECR-Net · RESIDE | PSNR: 33.85 -> 35.50; 1.65dB | Helped |
| add DFE to the base network with mixup (ablation) | AECR-Net · RESIDE | PSNR: 34.04 -> 36.20; 2.16dB | Helped |
| add positive-only CR to the AE-like network (ablation) | AECR-Net · RESIDE | PSNR: 36.20 -> 36.46; 0.26dB | Helped |
| include negative hazy samples in CR rather than using positive-only CR (ablation) | AECR-Net · RESIDE | PSNR: 36.46 -> 37.17 | Helped |
| deploy DFE after deeper layers rather than shallow layers (ablation) | AECR-Net · RESIDE | PSNR | Helped |
| increase negative samples from 1 to 10 while keeping positive samples at 1 (ablation) | AECR-Net CR · RESIDE | PSNR: 37.17 -> 37.41 | Helped |
| increase positive samples from 1 to 10 while keeping negative samples at 1 (ablation) | AECR-Net CR · RESIDE | PSNR: 37.17 -> 35.61 | Hurt |
| increase both positive and negative samples from 1 to 10 (ablation) | AECR-Net CR · RESIDE | PSNR: 37.17 -> 35.65 | Hurt |
| add CR to GridDehazeNet | GridDehazeNet · SOTS | PSNR: +0.83 | Helped |
Closest papers by meaning
- ODCR: Orthogonal Decoupling Contrastive Regularization for Unpaired Image Dehazing (CVPR 2024 · similarity 0.88)
- Contrastive Learning for Low-Light Raw Denoising (Student Abstract) (AAAI 2024 · similarity 0.87)
- Dehaze-RetinexGAN: Real-World Image Dehazing via Retinex-based Generative Adversarial Network (AAAI 2025 · similarity 0.87)
- Learning Hazing to Dehazing: Towards Realistic Haze Generation for Real-World Image Dehazing (CVPR 2025 · similarity 0.87)
- Exploiting Diffusion Prior for Real-World Image Dehazing with Unpaired Training (AAAI 2025 · similarity 0.87)
- PromptHaze: Prompting Real-world Dehazing via Depth Anything Model (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.