Boosting Diffusion Guidance via Learning Degradation-Aware Models for Blind Super Resolution
This work addresses the challenge of blind super-resolution for image processing, where degradation kernels are unknown, representing an incremental improvement by adapting existing diffusion guidance techniques.
The paper tackles the problem of generating high-resolution images with both high-frequency detail and fidelity in blind super-resolution by incorporating degradation-aware models into diffusion guidance, eliminating the need for known degradation kernels. The proposed method, DADiff, achieves superior performance over state-of-the-art methods on blind SR benchmarks.
Recently, diffusion-based blind super-resolution (SR) methods have shown great ability to generate high-resolution images with abundant high-frequency detail, but the detail is often achieved at the expense of fidelity. Meanwhile, another line of research focusing on rectifying the reverse process of diffusion models (i.e., diffusion guidance), has demonstrated the power to generate high-fidelity results for non-blind SR. However, these methods rely on known degradation kernels, making them difficult to apply to blind SR. To address these issues, we present DADiff in this paper. DADiff incorporates degradation-aware models into the diffusion guidance framework, eliminating the need to know degradation kernels. Additionally, we propose two novel techniques: input perturbation and guidance scalar, to further improve our performance. Extensive experimental results show that our proposed method has superior performance over state-of-the-art methods on blind SR benchmarks.