BlurDM: A Blur Diffusion Model for Image Deblurring

TL;DR

BlurDM introduces a novel diffusion-based approach that explicitly models the blur formation process for improved image deblurring, achieving significant performance gains across multiple benchmarks.

Contribution

It presents a dual-diffusion framework that integrates blur modeling into diffusion models and operates in latent space for efficient deblurring.

Findings

Outperforms existing deblurring methods on four benchmarks

Effectively models motion blur as a continuous process

Enhances deblurring by joint denoising and deblurring

Abstract

Diffusion models show promise for dynamic scene deblurring; however, existing studies often fail to leverage the intrinsic nature of the blurring process within diffusion models, limiting their full potential. To address it, we present a Blur Diffusion Model (BlurDM), which seamlessly integrates the blur formation process into diffusion for image deblurring. Observing that motion blur stems from continuous exposure, BlurDM implicitly models the blur formation process through a dual-diffusion forward scheme, diffusing both noise and blur onto a sharp image. During the reverse generation process, we derive a dual denoising and deblurring formulation, enabling BlurDM to recover the sharp image by simultaneously denoising and deblurring, given pure Gaussian noise conditioned on the blurred image as input. Additionally, to efficiently integrate BlurDM into deblurring networks, we perform BlurDM in the latent space, forming a flexible prior generation network for deblurring. Extensive experiments demonstrate that BlurDM significantly and consistently enhances existing deblurring methods on four benchmark datasets. The project page is available at https://jin-ting-he.github.io/BlurDM/.