DaAIR

Efficient Degradation-aware Any Image Restoration

Eduard Zamfir¹, Zongwei Wu¹, Nancy Mehta¹, Danda Pani Paudel^2,3, Yulun Zhang⁴ and Radu Timofte¹

¹ University of Würzburg, Germany - ² INSAIT Sofia University, Bulgaria - ³ ETH Zürich, Switzerland - ⁴ Shanghai Jiao Tong University, China

Latest

• 05/24/2024: Repository is created.

Method

Abstract

Reconstructing missing details from degraded low-quality inputs poses a significant challenge. Recent progress in image restoration has demonstrated the efficacy of learning large models capable of addressing various degradations simultaneously. Nonetheless, these approaches introduce considerable computational overhead and complex learning paradigms, limiting their practical utility. In response, we propose DaAIR, an efficient All-in-One image restorer employing a Degradation-aware Learner (DaLe) in the low-rank regime to collaboratively mine shared aspects and subtle nuances across diverse degradations, generating a degradation-aware embedding. By dynamically allocating model capacity to input degradations, we realize an efficient restorer integrating holistic and specific learning within a unified model. Furthermore, DaAIR introduces a cost-efficient parameter update mechanism that enhances degradation awareness while maintaining computational efficiency. Extensive comparisons across five image degradations demonstrate that our DaAIR outperforms both state-of-the-art All-in-One models and degradation-specific counterparts, affirming our efficacy and practicality.

Citation

If you find our work helpful, please consider citing the following paper and/or ⭐ the repo.

@misc{zamfir2024efficient,
      title={Efficient Degradation-aware Any Image Restoration}, 
      author={Eduard Zamfir and Zongwei Wu and Nancy Mehta and Danda Pani Paudel and Yulun Zhang and Radu Timofte},
      year={2024},
      eprint={2405.15475},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}

Acknowledgements

This code is built on PromptIR and AirNet.