arXiv 2410.093062025-06-23cs.CV

TD-Paint：通过时间感知像素条件化实现更快的扩散图像修复

TD-Paint: Faster Diffusion Inpainting Through Time Aware Pixel Conditioning

INSA Rouen Normandie, LITIS UR 4108（法国 Rouen 规划大学）；University of York（约克大学）；East China Normal University（华东师范大学）；Université Rouen Normandie, LITIS UR 4108（法国 Rouen 规划大学）；LIVIA, Dept. of Systems Engineering, ETS Montreal（蒙特利尔工程学院）；Université Caen Normandie, CNRS, GREYC UMR6072（法国 Caen 规划大学）

Tsiry Mayet, Pourya Shamsolmoali, Simon Bernard, Eric Granger, Romain Hérault, Clement Chatelain

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

提出TD-Paint方法，通过在像素级建模可变噪声水平，在扩散过程早期嵌入已知像素信息，引导生成朝向目标流形，无需架构修改即显著加速采样并保持图像质量。

摘要：

扩散模型已成为高度有效的图像修复技术，然而，它们仍受限于缓慢的采样率。尽管最近的进展提高了生成质量，但也增加了采样时间，从而限制了在现实世界应用中的可扩展性。我们研究了基于扩散的修复模型的生成采样过程，并观察到这些模型在初始采样步骤中极少使用输入条件。结果，采样轨迹偏离了数据流形，需要复杂的同步机制来重新对齐生成过程。为了解决这个问题，我们提出了时间感知扩散绘制（TD-Paint），这是一种通过在像素级别建模可变噪声水平来调整扩散过程的新方法。该技术允许模型从一开始就高效利用已知像素值，引导生成过程朝向目标流形。通过在扩散过程早期嵌入这些信息，TD-Paint显著加速了采样而不损害图像质量。与需要专用架构或昂贵生成循环的传统基于扩散的修复模型不同，TD-Paint无需架构修改即可实现更快的采样时间。在三个数据集上的实验结果表明，TD-Paint优于最先进的扩散模型，同时保持了较低的复杂度。

英文摘要：

Diffusion models have emerged as highly effective techniques for inpainting, however, they remain constrained by slow sampling rates. While recent advances have enhanced generation quality, they have also increased sampling time, thereby limiting scalability in real-world applications. We investigate the generative sampling process of diffusion-based inpainting models and observe that these models make minimal use of the input condition during the initial sampling steps. As a result, the sampling trajectory deviates from the data manifold, requiring complex synchronization mechanisms to realign the generation process. To address this, we propose Time-aware Diffusion Paint (TD-Paint), a novel approach that adapts the diffusion process by modeling variable noise levels at the pixel level. This technique allows the model to efficiently use known pixel values from the start, guiding the generation process toward the target manifold. By embedding this information early in the diffusion process, TD-Paint significantly accelerates sampling without compromising image quality. Unlike conventional diffusion-based inpainting models, which require a dedicated architecture or an expensive generation loop, TD-Paint achieves faster sampling times without architectural modifications. Experimental results across three datasets show that TD-Paint outperforms state-of-the-art diffusion models while maintaining lower complexity.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。