Research map / Image generation / Autoregressive and latent generation
Does Haze Removal Help CNN-based Image Classification?
Yanting Pei, Yaping Huang, Qi Zou, Yuhang Lu, Song Wang
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 |
|---|---|---|---|
| apply haze (beta=5) to test images when classifier trained on clear images | AlexNet classifier · Caltech-256 | accuracy (%): 71.7 -> 21.7 | Hurt |
| add MSCNN dehazing to hazy test images (classifier trained on clear) | AlexNet classifier · Caltech-256 beta=5 | accuracy (%): 21.7 -> 42.5 | Improved |
| train classifier on hazy images of same level instead of clear images | AlexNet classifier · Caltech-256 beta=5 | accuracy (%): 42.5 -> 51.9 | Improved |
| apply dehazing to both training and testing images | AlexNet, VGGNet-16, ResNet-50 · Caltech-256 all haze levels and Haze-20 | accuracy | No clear effect |
| apply dehazing for sparse coding classifier | sparse coding classifier · Caltech-256 and Haze-20 | accuracy | No clear effect |
| apply dehazing when training on mixed-level hazy images | CNN classifiers · Caltech-256 mixed haze levels and Haze-20 + HazeClear-20 | accuracy | No clear effect |
| select dehazed images judged subjectively better rather than worse | VGGNet classifier · Caltech-256 beta=3 dehazed subsets | accuracy (%) | No clear effect |
| rank dehazing methods by PSNR vs classification accuracy | dehazing metrics · Caltech-256 beta=1..5 | rank correlation: -0.3095 -> 0.3571 | Mixed |
| rank dehazing methods by SSIM vs classification accuracy | dehazing metrics · Caltech-256 beta=1..5 | rank correlation: -0.5238 -> 0.6905 | Mixed |
Closest papers by meaning
- Advancing Real-World Image Dehazing: Perspective, Modules, and Training (TPAMI 2024 · similarity 0.89)
- Fully Zero-Shot Image Dehazing (ICML 2026 · similarity 0.89)
- Detection-Friendly Dehazing: Object Detection in Real-World Hazy Scenes (TPAMI 2023 · similarity 0.88)
- Learning Hazing to Dehazing: Towards Realistic Haze Generation for Real-World Image Dehazing (CVPR 2025 · similarity 0.87)
- Tokenize Image Patches: Global Context Fusion for Effective Haze Removal in Large Images (CVPR 2025 · similarity 0.86)
- Real-world Image Dehazing with Coherence-based Pseudo Labeling and Cooperative Unfolding Network (NeurIPS 2024 · 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.
