DR2: Diffusion-based Robust Degradation Remover for Blind Face Restoration
This work improves face restoration robustness for real-world applications, but it is incremental as it builds on existing diffusion models and enhancement modules.
The paper tackles the problem of blind face restoration by addressing the gap between assumed and actual degradation in real-world images, proposing DR2 to transform degraded images into a degradation-invariant coarse prediction and then enhance it, resulting in outperforming state-of-the-art methods on heavily degraded datasets.
Blind face restoration usually synthesizes degraded low-quality data with a pre-defined degradation model for training, while more complex cases could happen in the real world. This gap between the assumed and actual degradation hurts the restoration performance where artifacts are often observed in the output. However, it is expensive and infeasible to include every type of degradation to cover real-world cases in the training data. To tackle this robustness issue, we propose Diffusion-based Robust Degradation Remover (DR2) to first transform the degraded image to a coarse but degradation-invariant prediction, then employ an enhancement module to restore the coarse prediction to a high-quality image. By leveraging a well-performing denoising diffusion probabilistic model, our DR2 diffuses input images to a noisy status where various types of degradation give way to Gaussian noise, and then captures semantic information through iterative denoising steps. As a result, DR2 is robust against common degradation (e.g. blur, resize, noise and compression) and compatible with different designs of enhancement modules. Experiments in various settings show that our framework outperforms state-of-the-art methods on heavily degraded synthetic and real-world datasets.