Residual Diffusion Implicit Models

2026-09-26

AI Summary

This study addresses the limitations of diffusion models in solving inverse problems, including noise initialization mismatch, hallucination artifacts in bridging models, and the lack of variance modeling in consistency models. To this end, we propose a Residual Diffusion Implicit Model that explicitly models the residual between high- and low-quality images to align with real degradation processes. The method introduces a non-Markovian implicit reverse sampler for single-step reconstruction and designs a controllable variance mechanism to balance fidelity and diversity, jointly optimized via perceptual loss. Experimental results demonstrate that the proposed model outperforms state-of-the-art methods in denoising and super-resolution tasks across metrics such as PSNR, significantly suppressing artifact generation while requiring minimal sampling steps.

Abstract

Diffusion models achieve state-of-the-art results across multiple tasks. However, in inverse problems, standard initialization from pure Gaussian noise misaligns the generative process with real-world degradations. More recent methods such as diffusion bridges impose strict endpoint constraints and often require long reverse processes that are prone to hallucinations. Alternative consistency models provide noise-invariant, one-step mappings but lack inherent variance modeling and can degrade under severe corruption. Hence, residual diffusion implicit models (RDIMs) are proposed, constituting a generalized framework that explicitly models the residuals between high-quality (HQ) and low-quality (LQ) images, aligning the forward process with the actual degradation. A non-Markovian implicit reverse sampler is derived, which can skip intermediate timesteps, enabling accurate few-step or even single-step reconstruction, while mitigating the hallucinations inherent to long diffusion chains. RDIM also introduces a controllable variance mechanism that interpolates between deterministic and stochastic sampling, balancing fidelity and diversity. Furthermore, it enables the straightforward use of perceptual losses, when needed. Experiments on denoising and super-resolution benchmarks demonstrate that RDIMs consistently outperforms the state of the art, including bridge and consistency models, in terms of PSNR, SSIM, and LPIPS, reducing hallucinations while requiring only a few sampling steps (often just one). The results position RDIMs as an efficient solution for a broad range of image restoration tasks.

Problem

• Inverse Problems
• Image Restoration
• Diffusion Models
• Hallucination
• Degradation Alignment

Innovation

• Residual Diffusion Implicit Models
• Non-Markovian Reverse Sampler
• Controllable Variance Mechanism
• Inverse Problems
• Image Restoration