图像编辑模型中的早期语义接地用于零样本指引用图像分割
Early Semantic Grounding in Image Editing Models for Zero-Shot Referring Image Segmentation
AI总结：
本文探讨利用图像编辑模型的隐含语义接地能力进行零样本指引用图像分割，通过分析发现早期去噪步骤即可实现前景背景分离，提出无需训练的框架利用预训练模型的中间表示进行分割。
摘要：
基于指令的图像编辑（IIE）模型最近展示了根据自然语言指令修改特定图像区域的强大能力，这隐含地要求识别应应用编辑的区域。本文研究是否可以利用这种隐含的接地能力进行零样本指引用图像分割（RIS），该任务需要对由自然语言描述的对象进行像素级定位。通过系统分析，发现这些模型在最早去噪时间步内部表示中出现强前景-背景分离，远早于任何可见图像变换发生。基于此见解，我们提出一个无需训练的框架，通过利用预训练图像编辑模型的中间表示来重新利用它们进行RIS。我们的方法将定位分解为两个互补组件：基于注意力的空间先验估计关注区域，以及基于特征的语义辨别确定分割对象。通过利用特征空间分离性，该框架仅使用单步去噪即可生成准确的分割掩码，无需完整图像合成。在RefCOCO、RefCOCO+和RefCOCOg上的大量实验表明，我们的方法在现有零样本基线中表现更优。
英文摘要：
Instruction-based image editing (IIE) models have recently demonstrated strong capability in modifying specific image regions according to natural language instructions, which implicitly requires identifying where an edit should be applied. This indicates that such models inherently perform language-conditioned visual semantic grounding. In this work, we investigate whether this implicit grounding can be leveraged for zero-shot referring image segmentation (RIS), a task that requires pixel-level localization of objects described by natural language expressions. Through systematic analysis, we reveal that strong foreground-background separability emerges in the internal representations of these models at the earliest denoising timestep, well before any visible image transformation occurs. Building on this insight, we propose a training-free framework that repurposes pretrained image editing models for RIS by exploiting their intermediate representations. Our approach decomposes localization into two complementary components: attention-based spatial priors that estimate where to focus, and feature-based semantic discrimination that determines what to segment. By leveraging feature-space separability, the framework produces accurate segmentation masks using only a single denoising step, without requiring full image synthesis. Extensive experiments on RefCOCO, RefCOCO+, and RefCOCOg demonstrate that our method achieves superior performance over existing zero-shot baselines.
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。
