MoEdit：关于多对象图像编辑中数量感知的学习
MoEdit: On Learning Quantity Perception for Multi-object Image Editing
AI总结：
本文提出无需辅助工具的 MoEdit 框架，通过 FeCom 与 QTTN 模块在 SD 上保持多对象数量一致，实现风格迁移、对象重塑和背景再生的 SOTA 编辑效果。
摘要：
多对象图像在增强现实、广告设计和医学成像等各种真实场景中十分普遍。对这些图像进行高效而精确的编辑，对上述应用至关重要。随着 Stable Diffusion（SD）的出现，高质量图像生成与编辑进入了一个新时代。然而，现有方法往往难以同时将每个对象视为独立个体，并将其作为整幅图像编辑的一部分来考虑，而这两点对于确保一致的数量感知都至关重要，因此会导致次优的感知性能。为应对这些挑战，我们提出了 MoEdit，这是一个无需辅助工具的多对象图像编辑框架。MoEdit 促进了风格迁移、对象重塑和背景再生方面的高质量多对象图像编辑，同时即使在对象数量很多时，也能确保输入与输出之间具有一致的数量感知。为实现这一点，我们引入了 Feature Compensation（FeCom，特征补偿）模块，该模块通过最小化对象之间的交错，确保每个对象属性的区分度和可分离性。此外，我们提出了 Quantity Attention（QTTN，数量注意力）模块，该模块在编辑过程中通过有效控制来感知并保持数量一致性，而不依赖辅助工具。通过利用 SD 模型，MoEdit 能够高质量地对输入中的特定概念进行定制化保留和修改。实验结果表明，我们的 MoEdit 在多对象图像编辑中取得了 State-Of-The-Art（SOTA）性能。数据和代码将在 https://github.com/Tear-kitty/MoEdit 提供。
英文摘要：
Multi-object images are prevalent in various real-world scenarios, including augmented reality, advertisement design, and medical imaging. Efficient and precise editing of these images is critical for these applications. With the advent of Stable Diffusion (SD), high-quality image generation and editing have entered a new era. However, existing methods often struggle to consider each object both individually and part of the whole image editing, both of which are crucial for ensuring consistent quantity perception, resulting in suboptimal perceptual performance. To address these challenges, we propose MoEdit, an auxiliary-free multi-object image editing framework. MoEdit facilitates high-quality multi-object image editing in terms of style transfer, object reinvention, and background regeneration, while ensuring consistent quantity perception between inputs and outputs, even with a large number of objects. To achieve this, we introduce the Feature Compensation (FeCom) module, which ensures the distinction and separability of each object attribute by minimizing the in-between interlacing. Additionally, we present the Quantity Attention (QTTN) module, which perceives and preserves quantity consistency by effective control in editing, without relying on auxiliary tools. By leveraging the SD model, MoEdit enables customized preservation and modification of specific concepts in inputs with high quality. Experimental results demonstrate that our MoEdit achieves State-Of-The-Art (SOTA) performance in multi-object image editing. Data and codes will be available at https://github.com/Tear-kitty/MoEdit.
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。
