arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2602.11850cs.CVcs.LG
收藏

为稳定修正流反向操作获取免费午餐

Free Lunch for Stabilizing Rectified Flow Inversion

Westlake University(西湖大学);Nanyang Technological University(南洋理工大学)

Chenru Wang, Beier Zhu, Chi Zhang

AI总结:

本文提出PMI和mimic-CFG方法,通过稳定速度场提升修正流反向操作的稳定性与重建质量,减少计算资源消耗。

摘要:

基于修正流(RF)的生成模型最近涌现出作为传统扩散模型的有力替代品,在各种任务中展示了最先进的性能。通过学习一个连续的速度场,将简单噪声转化为复杂数据,RF模型不仅能够实现高质量的生成,还支持无训练反向操作,这有助于下游任务如重建和编辑。然而,现有的反向方法,如 vanilla RF-based 反向,受到近似误差的困扰,这些误差在时间步之间累积,导致速度场不稳定,重建和编辑质量下降。为了解决这一挑战,我们提出了Proximal-Mean Inversion(PMI),一种无训练梯度校正方法,通过引导速度场朝向过去速度的运行平均值,受理论推导出的球形高斯约束。此外,我们引入了mimic-CFG,一种轻量的速度校正方案用于编辑任务,它在当前速度和其在历史平均上的投影之间进行插值,平衡编辑效果和结构一致性。在PIE-Bench上的大量实验表明,我们的方法显著提高了反向稳定性、图像重建质量和编辑保真度,同时减少了所需的神经功能评估次数。我们的方法在PIE-Bench上实现了最先进的性能,具有增强的效率和理论严谨性。

英文摘要:

Rectified-Flow (RF)-based generative models have recently emerged as strong alternatives to traditional diffusion models, demonstrating state-of-the-art performance across various tasks. By learning a continuous velocity field that transforms simple noise into complex data, RF-based models not only enable high-quality generation, but also support training-free inversion, which facilitates downstream tasks such as reconstruction and editing. However, existing inversion methods, such as vanilla RF-based inversion, suffer from approximation errors that accumulate across timesteps, leading to unstable velocity fields and degraded reconstruction and editing quality. To address this challenge, we propose Proximal-Mean Inversion (PMI), a training-free gradient correction method that stabilizes the velocity field by guiding it toward a running average of past velocities, constrained within a theoretically derived spherical Gaussian. Furthermore, we introduce mimic-CFG, a lightweight velocity correction scheme for editing tasks, which interpolates between the current velocity and its projection onto the historical average, balancing editing effectiveness and structural consistency. Extensive experiments on PIE-Bench demonstrate that our methods significantly improve inversion stability, image reconstruction quality, and editing fidelity, while reducing the required number of neural function evaluations. Our approach achieves state-of-the-art performance on the PIE-Bench with enhanced efficiency and theoretical soundness.

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。

↑