Seeing Through the Rain: Resolving High-Frequency Conflicts in Deraining and Super-Resolution via Diffusion Guidance

AAAI 2026

Wenjie Li1, Jinglei Shi2, Jin Han3, Heng Guo1, Zhanyu Ma1,
1Beijing University of Posts and Telecommunications
2NanKai University 3Tokyo University
Arxiv Code Supplementary

Abstract

Clean images are crucial for visual tasks such as small object detection, especially at high resolutions. However, real-world images are often degraded by adverse weather, and weather restoration methods may sacrifice high-frequency details critical for analyzing small objects. A natural solution is to apply super-resolution (SR) after weather removal to recover both clarity and fine structures. However, simply cascading restoration and SR struggle to bridge their inherent conflict: removal aims to remove high-frequency weather-induced noise, while SR aims to hallucinate high-frequency textures from existing details, leading to inconsistent restoration contents.

In this paper, we take deraining as a case study and propose DHGM, a Diffusion-based High-frequency Guided Model for generating clean and high-resolution images. DHGM integrates pre-trained diffusion priors with high-pass filters to simultaneously remove rain artifacts and enhance structural details. Extensive experiments demonstrate that DHGM achieves superior performance over existing methods, with lower costs.

Overview of Our Method

Overview: Our method utilizes our Media Remover (MR) and Texture Compensator (TC) to guide learned diffusion priors in latent spaces to complete high-frequency rain-induced media removal and high-frequency texture reconstruction.

Quantitative Results

Quantitative: Results of ours with sequentially performed deraining and SR methods, fine-tuned single deraining, all-in-one weather restoration, SR, and general restoration methods on deraining, deraining & raindrop removal, and raindrop removal test sets at scale of x2.

Qualitative Results

Qualitative: The sequential utilization of existing deraining methods and super-resolution methods or fine-tuned deraining methods makes it difficult to accurately recover buildings, vehicles, and pedestrians (top), hindering downstream tasks like depth map estimation and image segmentation (bottom).

BibTeX

@inproceedings{li2026seeing,
      title={Seeing Through the Rain: Resolving High-Frequency Conflicts in Deraining and Super-Resolution via Diffusion Guidance},
      author={Li, Wenjie and Shi, Jinglei and Han, Jin and Guo, Heng and Ma, Zhanyu},
      booktitle={AAAI},
      year={2026}
    }