Deblurring via Stochastic Refinement

TL;DR

This paper introduces a stochastic diffusion model framework for blind image deblurring that produces diverse, perceptually realistic reconstructions, outperforming existing methods in perceptual quality while maintaining competitive distortion metrics.

Contribution

It proposes a novel stochastic sampler that refines deterministic predictions, enabling diverse image reconstructions and improved perceptual quality in deblurring tasks.

Findings

Significant improvement in perceptual quality over state-of-the-art methods.

Efficient sampling compared to traditional diffusion models.

Competitive distortion metrics like PSNR achieved.

Abstract

Image deblurring is an ill-posed problem with multiple plausible solutions for a given input image. However, most existing methods produce a deterministic estimate of the clean image and are trained to minimize pixel-level distortion. These metrics are known to be poorly correlated with human perception, and often lead to unrealistic reconstructions. We present an alternative framework for blind deblurring based on conditional diffusion models. Unlike existing techniques, we train a stochastic sampler that refines the output of a deterministic predictor and is capable of producing a diverse set of plausible reconstructions for a given input. This leads to a significant improvement in perceptual quality over existing state-of-the-art methods across multiple standard benchmarks. Our predict-and-refine approach also enables much more efficient sampling compared to typical diffusion models. Combined with a carefully tuned network architecture and inference procedure, our method is competitive in terms of distortion metrics such as PSNR. These results show clear benefits of our diffusion-based method for deblurring and challenge the widely used strategy of producing a single, deterministic reconstruction.