NEAF：面向文本引导图像编辑中可泛化测试时优化的注意力融合自然图像编辑方法
NEAF: Natural Image Editing with Attention Fusion for Generalizable Test-time Optimization in Text-Guided Image Editing
提出可学习的 XA-Conductor 模块，在推理时对源分支、编辑分支与重建分支的交叉注意力图做自适应加权融合，无需微调即可用于任意 T2I 模型；做文本引导编辑、需要保持非编辑区域保真度的工作可直接复用这一三支反馈的注意力融合机制。
原文依据 a lightweight, learnable XA-Conductor module that dynamically identifies salient cross-attention contributions pertinent to the edit
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论基于扩散的文本到图像（T2I）模型已具备出色的生成能力，但如何在保持原图结构与感知保真度的前提下进行精确的文本引导图像编辑仍非易事。现有方法要么依赖大规模定制数据集重新训练，带来高昂的计算与数据整理成本；要么采用轻量微调策略，但仍需优化，且在细粒度或语义复杂的编辑中常常失效。我们提出 NEAF（Natural image Editing with Attention Fusion），一种新颖的零样本、通用、免调优框架，可适用于任意 T2I 模型，无需数据集整理或重新训练。NEAF 引入轻量可学习的 XA-Conductor 模块，动态识别与编辑相关的显著交叉注意力贡献。该模块优化一个权重向量，以协调来自源分支、编辑分支与重建分支的交叉注意力图的自适应融合。这一三支反馈优化策略确保用户指令被精确落实，同时严格保持静止区域的保真度。大量实验验证 NEAF 是一个灵活且通用的框架，在多种编辑任务上持续超越现有方法，在其它方法易失效的复杂非刚性编辑场景中表现尤为突出。
英文原文摘要 1399 字符
Diffusion-based text-to-image (T2I) models have enabled remarkable generative capabilities, yet precise text-based image editing that preserves the original's structural and perceptual fidelity remains non-trivial. Existing approaches either rely on retraining with large bespoke datasets, incurring significant computational and curation costs, or adopt lightweight fine-tuning strategies that still require optimization and often fail in fine-grained or semantically complex edits. We propose NEAF (Natural image Editing with Attention Fusion), a novel zero-shot, universal tuning-free framework for arbitrary T2I models, obviating the need for dataset curation or retraining. NEAF introduces a lightweight, learnable XA-Conductor module that dynamically identifies salient cross-attention contributions pertinent to the edit. This module optimizes a weight vector to orchestrate an adaptive fusion of cross-attention maps derived from the source, edited, and reconstruction branches. This triadic-feedback optimization strategy ensures the precise instantiation of user directives while rigorously preserving the fidelity of quiescent regions. Extensive experiments validate NEAF as a flexible and general framework that consistently surpasses existing methods across diverse editing tasks, demonstrating particular dominance in complex, non-rigid editing scenarios where other approaches falter.