扩散反演中的时间步重调度
Timestep Rescheduling in Diffusion Inversion
AI总结：
针对扩散反演中时间步选择影响反演精度的问题，提出一种基于全局重缩放和局部动态规划的非均匀时间步调度器，有效降低反演误差，提升图像重建与编辑性能。
摘要：
扩散反演将图像映射回扩散模型的高斯潜在空间，是图像重建和编辑的关键任务。虽然DDIM实现了快速确定性反演，但它固有地引入了累积为明显反演误差的偏差。现有方法通常通过求解不动点问题来解决这一问题，但很大程度上忽略了噪声调度器中扩散时间步的选择如何影响反演保真度。在这项工作中，我们揭示了扩散反演中的偏差尺度强烈依赖于时间步大小，并呈现出抛物线趋势，较大的误差集中在较小和较大的时间步。基于这一发现，我们提出了一种简单而有效的非均匀时间步调度器，该调度器集成了全局重缩放和基于局部动态规划的重调度，实现了计算资源的战略分配，从而最小化整体反演误差并保持更高的反演精度。我们的方法可作为现有反演技术的即插即用增强，无需额外参数或计算开销。通过大量实验，我们验证了集成我们的调度器能够持续提升现有反演方法的性能，在图像重建和编辑中取得更优结果。
英文摘要：
Diffusion inversion, which maps images back to the Gaussian latent space of a diffusion model, is a critical task for image reconstruction and editing. While DDIM enables fast deterministic inversion, it inherently introduces deviations that accumulate into noticeable inversion errors. Existing methods often address this by solving a fixed-point problem but largely overlook how the selection of the diffusion timestep in the noise scheduler influences inversion fidelity. In this work, we reveal that the deviation scale in diffusion inversion is strongly dependent on the timestep size, and exhibits a parabolic trend, with larger errors concentrated at both small and large timesteps. Based on this finding, we propose a simple yet effective nonuniform timestep scheduler that integrates a global rescaling with a local dynamic programming based rescheduling, enabling a strategic allocation of computational effort that minimizes the overall inversion error and preserves higher inversion accuracy. Our method serves as an off-the-shelf enhancement for existing inversion techniques and requires no extra parameters or computational overhead. Through extensive experiments, we verify that integrating our scheduler consistently boosts the performance of existing inversion methods, achieving superior results in image reconstruction and editing.
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。
