DirectEdit: 基于流的图像编辑的逐步骤精确反演
DirectEdit: Step-Level Accurate Inversion for Flow-Based Image Editing
AI总结：
提出DirectEdit方法，通过直接对齐前向路径消除反演过程中的累积漂移，实现精确重建和可靠特征共享，无需额外神经函数评估，在多种场景下优于现有方法。
摘要：
随着大规模预训练文本到图像（T2I）模型的最新进展，免训练的图像编辑方法已展现出显著成功。通常，这些方法通过反演过程向干净图像添加噪声，随后在前向过程中分别对重建路径和编辑路径进行去噪步骤。然而，由于重建路径使用来自不匹配时间步的噪声潜变量进行近似，现有方法不可避免地遭受累积漂移，这从根本上限制了重建保真度。为了解决这一挑战，我们系统地分析了流变换器中的反演过程，并提出了DirectEdit，一种简单而有效的编辑方法，无需引入额外的神经函数评估（NFE）即可消除固有的重建误差。与大多数试图纠正反演路径的先前工作不同，DirectEdit专注于直接对齐前向路径，从而实现精确重建和可靠的特征共享。此外，我们引入了一种基于注意力特征注入和多分支掩码引导噪声混合的保留机制，有效平衡了保真度和可编辑性。跨多种场景的大量实验表明，DirectEdit实现了高效准确的图像编辑，其优越性能优于最先进的方法。代码和示例可在 https://desongyang.github.io/Directedit 获取。
英文摘要：
With recent advancements in large-scale pre-trained text-to-image (T2I) models, training-free image editing methods have demonstrated remarkable success. Typically, these methods involve adding noise to a clean image via an inversion process, followed by separate denoising steps for the reconstruction and editing paths during the forward process. However, since the reconstruction path is approximated using noisy latents from mismatched timesteps, existing methods inevitably suffer from accumulated drift, which fundamentally limits reconstruction fidelity. To address this challenge, we systematically analyze the inversion process within the flow transformer and propose DirectEdit, a simple yet effective editing method that eliminates the inherent reconstruction error without introducing additional neural function evaluations (NFEs). Unlike most prior works that attempt to rectify the inversion path, DirectEdit focuses on directly aligning the forward paths, enabling precise reconstruction and reliable feature sharing. Furthermore, we introduce a preservation mechanism based on attention feature injection and multi-branch mask-guided noise blending, which effectively balances fidelity and editability. Extensive experiments across diverse scenarios demonstrate that DirectEdit achieves efficient and accurate image editing, delivering superior performance that outperforms state-of-the-art methods. Code and examples are available at https://desongyang.github.io/Directedit.
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。
