残差扩散隐式模型
Residual Diffusion Implicit Models
浏览论文内容
中文总结 AI 辅助
针对扩散模型在逆问题中的初始化错位和长链幻觉问题，提出残差扩散隐式模型（RDIM），显式建模HQ-LQ残差，通过非马尔可夫隐式采样实现少步重建，并引入可控方差机制，在去噪和超分辨率基准上超越现有方法。
中文摘要 AI 辅助
扩散模型在多个任务上取得了最先进的结果。然而，在逆问题中，从纯高斯噪声的标准初始化会使生成过程与真实世界的退化错位。更近期的方法如扩散桥施加严格的端点约束，并且通常需要较长的反向过程，这容易产生幻觉。另一种一致性模型提供噪声不变的单步映射，但缺乏固有的方差建模，并且在严重损坏下可能退化。因此，提出了残差扩散隐式模型（RDIMs），构成一个广义框架，显式地建模高质量（HQ）和低质量（LQ）图像之间的残差，使前向过程与实际退化对齐。推导了一种非马尔可夫隐式反向采样器，可以跳过中间时间步，实现准确的少步甚至单步重建，同时减轻长扩散链固有的幻觉。RDIM还引入了一种可控方差机制，在确定性和随机采样之间插值，平衡保真度和多样性。此外，它使得在需要时能够直接使用感知损失。在去噪和超分辨率基准上的实验表明，RDIMs在PSNR、SSIM和LPIPS方面持续优于最先进的方法，包括桥模型和一致性模型，减少了幻觉，同时只需要少量采样步骤（通常仅一步）。结果将RDIMs定位为广泛图像恢复任务的高效解决方案。
英文摘要
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.
