🤖 AI Summary
This work addresses a fundamental limitation in conventional diffusion models for image deblurring, which erroneously model blur degradation as additive noise while neglecting its inherently convolutional nature. To overcome this, the authors propose ConvDiff, a novel framework that aligns the forward diffusion process with the physical convolutional mechanism of blur. By constructing a degradation trajectory consistent with real-world blur through frequency-domain decomposition and integrating Gaussian blur modeling with diffusion inversion algorithms, ConvDiff transcends the restrictive additive noise assumption. This approach yields deblurring results that better adhere to physical principles and establishes a scalable diffusion-based paradigm applicable to diverse blur types.
📝 Abstract
In recent times, diffusion models have achieved remarkable performance in image restoration tasks. Their core mechanism relies on the restricted presumption of degradation prior to the additive noise operation. However, the blur model, one of the most widely studied degradation formulations, violates this assumption, as it is inherently based on convolution rather than addition. In this paper, we introduce ConvDiff, a novel diffusion based framework that substitutes the additive operation with convolution for the task of image deblurring. In the forward process, we construct a meaningful trajectory from the clean image to its blurred counterpart by exploiting the frequency domain characteristics of convolution, rather than progressively corrupting the image with additive noise. While the current work instantiates this framework for Gaussian blur, where frequency-domain decomposition yields closed-form and physically valid intermediate states, the underlying principle of constructing degradation trajectories from the blur operator extends naturally to other blur families. This formulation bridges the gap between the mathematical principles of blurring and the iterative design of diffusion-based restoration algorithms, enabling more physically grounded and effective image restoration models.