Learning Hazing to Dehazing: Towards Realistic Haze Generation for Real-World Image Dehazing
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866912292779589632 |
|---|---|
| author | Wang, Ruiyi Zheng, Yushuo Zhang, Zicheng Li, Chunyi Liu, Shuaicheng Zhai, Guangtao Liu, Xiaohong |
| author_facet | Wang, Ruiyi Zheng, Yushuo Zhang, Zicheng Li, Chunyi Liu, Shuaicheng Zhai, Guangtao Liu, Xiaohong |
| contents | Existing real-world image dehazing methods primarily attempt to fine-tune pre-trained models or adapt their inference procedures, thus heavily relying on the pre-trained models and associated training data. Moreover, restoring heavily distorted information under dense haze requires generative diffusion models, whose potential in dehazing remains underutilized partly due to their lengthy sampling processes. To address these limitations, we introduce a novel hazing-dehazing pipeline consisting of a Realistic Hazy Image Generation framework (HazeGen) and a Diffusion-based Dehazing framework (DiffDehaze). Specifically, HazeGen harnesses robust generative diffusion priors of real-world hazy images embedded in a pre-trained text-to-image diffusion model. By employing specialized hybrid training and blended sampling strategies, HazeGen produces realistic and diverse hazy images as high-quality training data for DiffDehaze. To alleviate the inefficiency and fidelity concerns associated with diffusion-based methods, DiffDehaze adopts an Accelerated Fidelity-Preserving Sampling process (AccSamp). The core of AccSamp is the Tiled Statistical Alignment Operation (AlignOp), which can provide a clean and faithful dehazing estimate within a small fraction of sampling steps to reduce complexity and enable effective fidelity guidance. Extensive experiments demonstrate the superior dehazing performance and visual quality of our approach over existing methods. The code is available at https://github.com/ruiyi-w/Learning-Hazing-to-Dehazing. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_19262 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Learning Hazing to Dehazing: Towards Realistic Haze Generation for Real-World Image Dehazing Wang, Ruiyi Zheng, Yushuo Zhang, Zicheng Li, Chunyi Liu, Shuaicheng Zhai, Guangtao Liu, Xiaohong Computer Vision and Pattern Recognition Existing real-world image dehazing methods primarily attempt to fine-tune pre-trained models or adapt their inference procedures, thus heavily relying on the pre-trained models and associated training data. Moreover, restoring heavily distorted information under dense haze requires generative diffusion models, whose potential in dehazing remains underutilized partly due to their lengthy sampling processes. To address these limitations, we introduce a novel hazing-dehazing pipeline consisting of a Realistic Hazy Image Generation framework (HazeGen) and a Diffusion-based Dehazing framework (DiffDehaze). Specifically, HazeGen harnesses robust generative diffusion priors of real-world hazy images embedded in a pre-trained text-to-image diffusion model. By employing specialized hybrid training and blended sampling strategies, HazeGen produces realistic and diverse hazy images as high-quality training data for DiffDehaze. To alleviate the inefficiency and fidelity concerns associated with diffusion-based methods, DiffDehaze adopts an Accelerated Fidelity-Preserving Sampling process (AccSamp). The core of AccSamp is the Tiled Statistical Alignment Operation (AlignOp), which can provide a clean and faithful dehazing estimate within a small fraction of sampling steps to reduce complexity and enable effective fidelity guidance. Extensive experiments demonstrate the superior dehazing performance and visual quality of our approach over existing methods. The code is available at https://github.com/ruiyi-w/Learning-Hazing-to-Dehazing. |
| title | Learning Hazing to Dehazing: Towards Realistic Haze Generation for Real-World Image Dehazing |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2503.19262 |