EEdit：重新思考高效图像编辑中的空间与时间冗余
EEdit: Rethinking the Spatial and Temporal Redundancy for Efficient Image Editing
AI总结：
EEdit框架通过空间局部性缓存和令牌索引预处理解决空间冗余，通过反演步骤跳过解决时间冗余，实现平均2.46倍加速且性能不降。
摘要：
基于反演的图像编辑正迅速获得关注，但其计算开销巨大，阻碍了其在实时交互场景中的应用。本文重新思考了基于反演的图像编辑中冗余存在于空间和时间两个维度，例如未编辑区域中的不必要计算以及反演过程中的冗余。为应对这些挑战，我们提出了一个实用框架，名为EEdit，以实现高效图像编辑。具体而言，我们引入了三种技术逐一解决这些问题。针对空间冗余，我们引入了空间局部性缓存，用于计算编辑区域及其邻近区域，同时跳过未编辑区域，并设计了令牌索引预处理以进一步加速缓存。针对时间冗余，我们提出了反演步骤跳过，以重用潜在表示实现高效编辑。我们的实验表明，在包括提示引导图像编辑、拖拽和图像合成在内的广泛编辑任务中，平均加速2.46倍，且性能无下降。我们的代码可在https://github.com/yuriYanZeXuan/EEdit获取。
英文摘要：
Inversion-based image editing is rapidly gaining momentum while suffering from significant computation overhead, hindering its application in real-time interactive scenarios. In this paper, we rethink that the redundancy in inversion-based image editing exists in both the spatial and temporal dimensions, such as the unnecessary computation in unedited regions and the redundancy in the inversion progress. To tackle these challenges, we propose a practical framework, named EEdit, to achieve efficient image editing. Specifically, we introduce three techniques to solve them one by one. For spatial redundancy, spatial locality caching is introduced to compute the edited region and its neighboring regions while skipping the unedited regions, and token indexing preprocessing is designed to further accelerate the caching. For temporal redundancy, inversion step skipping is proposed to reuse the latent for efficient editing. Our experiments demonstrate an average of 2.46 $\times$ acceleration without performance drop in a wide range of editing tasks including prompt-guided image editing, dragging and image composition. Our codes are available at https://github.com/yuriYanZeXuan/EEdit
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。
