Research map / Label-efficient and robust learning / Recognition and long-tail
Uncertainty-Driven Dehazing Network
Ming Hong, Jianzhuang Liu, Cuihua Li, Yanyun Qu
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 |
|---|---|---|---|
| apply UDN rather than AECR-Net | SOTS indoor dehazing · SOTS indoor | PSNR: 37.17 -> 38.62 | Improved |
| apply UDN rather than FFA-Net | SOTS outdoor dehazing · SOTS outdoor | PSNR: 33.38 -> 34.92 | Improved |
| apply UDN rather than KDDN | Middlebury dehazing · Middlebury | PSNR: 17.27 -> 18.43 | Improved |
| apply UDN rather than KDDN | O-HAZE dehazing · O-HAZE | PSNR: 25.455 -> 25.412 | Hurt |
| add uncertainty estimation blocks (ablation) | base dehazing network · SOTS indoor | PSNR: 35.53 -> 36.6 | Improved |
| add uncertainty-gated feature aggregation (ablation) | base+UEB · SOTS indoor | PSNR: 36.6 -> 36.72 | Improved |
| add uncertainty-aware feature modulation (ablation) | base+UEB+Gate · SOTS indoor | PSNR: 36.72 -> 38.21 | Improved |
| add self-distillation loss LSD (ablation) | base+UEB+Gate+UFM · SOTS indoor | PSNR: 38.21 -> 38.39 | Improved |
| use uncertainty-driven self-distillation loss LUSD instead of LSD (ablation) | base+UEB+Gate+UFM · SOTS indoor | PSNR: 38.39 -> 38.62 | Improved |
| use fused aleatoric and epistemic uncertainty instead of epistemic uncertainty only (ablation) | base dehazing network · SOTS indoor | PSNR: 35.84 -> 36.6 | Improved |
| use fused aleatoric and epistemic uncertainty instead of aleatoric uncertainty only (ablation) | base dehazing network · SOTS indoor | PSNR: 36.23 -> 36.6 | Improved |
| use fused uncertainty instead of epistemic uncertainty only (ablation) | UDN · SOTS indoor | PSNR: 37.89 -> 38.62 | Improved |
Closest papers by meaning
- Proximal Dehaze-Net: A Prior Learning-Based Deep Network for Single Image Dehazing (ECCV 2018 · similarity 0.88)
- Advancing Real-World Image Dehazing: Perspective, Modules, and Training (TPAMI 2024 · similarity 0.87)
- Deep Deterministic Uncertainty: A New Simple Baseline (CVPR 2023 · similarity 0.87)
- Learning Deep Priors for Image Dehazing (ICCV 2019 · similarity 0.87)
- SURE: SUrvey REcipes for building reliable and robust deep networks (CVPR 2024 · similarity 0.85)
- Learning Hazing to Dehazing: Towards Realistic Haze Generation for Real-World Image Dehazing (CVPR 2025 · similarity 0.85)
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.
