arXivDaily arXiv每日学术速递 周一至周五更新
​通知:联邦节假日调休:哥伦布日 / 原住民日(Columbus Day / Indigenous Peoples' Day),今日(13日)无更新。
arXiv 2403.12585cs.CV

LASPA:用于快速免训练单图像编辑的潜在空间对齐

LASPA: Latent Spatial Alignment for Fast Training-free Single Image Editing

NCAI(国家人工智能研究中心);KAUST(阿卜杜拉国王科技大学)

Yazeed Alharbi, Peter Wonka

AI总结:

本文提出一种基于潜在空间对齐(LASPA)的免训练扩散模型图像编辑方法,无需微调即可高效保留图像细节,实现语义连贯编辑,速度更快且存储需求更低,在用户研究中获得62-71%的偏好率。

摘要:

我们提出了一种使用扩散模型对真实图像进行文本编辑的新型免训练方法。与以往依赖计算成本高昂的微调方法不同,我们的方法利用LAtent SPatial Alignment(LASPA,潜在空间对齐)来高效保留图像细节。我们展示了如何使用参考图像对扩散过程进行空间引导,从而实现语义连贯的编辑。这消除了复杂优化的需求,避免了昂贵的模型微调,与以往方法相比,编辑速度显著提升。此外,该方法避免了大型微调模型所需的存储开销。这些优势使其特别适合移动设备上的编辑以及对响应时间要求严格的应用。尽管简单快速,该方法在用户研究中获得了62-71%的偏好率,并且在基于模型的编辑强度和图像保留评分上表现显著更优。

英文摘要:

We present a novel, training-free approach for textual editing of real images using diffusion models. Unlike prior methods that rely on computationally expensive finetuning, our approach leverages LAtent SPatial Alignment (LASPA) to efficiently preserve image details. We demonstrate how the diffusion process is amenable to spatial guidance using a reference image, leading to semantically coherent edits. This eliminates the need for complex optimization and costly model finetuning, resulting in significantly faster editing compared to previous methods. Additionally, our method avoids the storage requirements associated with large finetuned models. These advantages make our approach particularly well-suited for editing on mobile devices and applications demanding rapid response times. While simple and fast, our method achieves 62-71\% preference in a user-study and significantly better model-based editing strength and image preservation scores.

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。

↑