Dynamic Time Step Prediction in Inverse Heat Dissipation for Blur-Like Image Restoration Tasks

2026-10-02

AI Summary

This study addresses the limitations of diffusion models in blurry image restoration, specifically the use of fixed inversion steps and the mismatch between Gaussian noise and real-world degradations. To this end, we propose a novel diffusion framework based on inverse heat dissipation. Methodologically, the conventional Gaussian noising process is replaced by an inverse heat dissipation mechanism, enabling the forward process to physically align with actual image degradation. Furthermore, a learnable dynamic time prediction module is designed to adaptively estimate the optimal starting step for inversion, while a DDIM-variant architecture is employed to accelerate the solving process. Experimental results demonstrate that the proposed method achieves state-of-the-art performance on standard benchmarks and exhibits superior generalization capability across multiple restoration tasks.

Abstract

When using diffusion models to target image restoration problems, diffusion inversion is typically employed to retain relevant image information from the degraded images. Instead of inverting back to the initial time step (i.e., T), many methods invert to a pre-determined intermediate time step, in order to better preserve information from degraded source images. However, a pre-determined time step for inversion is not ideal for reconstruction, as a severely degraded image requires an earlier starting time step than a mildly degraded one. In addition, DDIM-based models corrupt the original signal by adding Gaussian noise, which can be mismatched to the nature of blur-like degradations, such as blur, haze, and low-light. To address these problems, we propose two solutions: (1) we adopt an alternative diffusion process, called the Inverse Heat Dissipation Model, that diffuses the input image by gradually blurring a data point (2) we propose to implement a time predictor to estimate the starting time step for the inversion, with the model learning to adapt to the degradation severity. Extensive experiments on standard benchmarks show that our method achieves state-of-the-art performance in both quantitative and qualitative evaluations, with excellent generalization to many restoration tasks.

Problem

• image restoration
• diffusion inversion
• dynamic time step
• blur-like degradation
• inverse heat dissipation

Innovation

• Inverse Heat Dissipation Model
• Dynamic Time Step Prediction
• Diffusion Inversion
• Image Restoration
• Blur-like Degradations