发表机构
The University of Hong Kong; ByteDance Seed; Zhejiang University（香港大学; 字节跳动 Seed; 浙江大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究提出可泛化框架FreeMatching，结合生成与语义基础表征及异构监督，提升IEG图像对的对应质量，可作为身份保留的定量指标。
AI 中文摘要
密集对应匹配长期以来受限于简化的时空先验，如平滑运动和刚性几何。这些假设在经典任务中有效，但在图像编辑和参考引导生成（IEG）中会失效，此类变换可在保留视觉身份的同时打破物理连续性。为在这类变换间建立身份保留的对应关系，我们提出FreeMatching，一种结合生成式与语义基础表征的可泛化框架，利用经典数据集、跟踪视频和合成场景的异构监督。教师引导的迭代细化无需密集对应标注即可进一步提升IEG中的对应质量。实验表明，单个FreeMatching模型可显著提升具有挑战性的IEG图像对的对应质量，同时在经典基准上保持竞争力。此外，我们证明其可作为评估身份保留的定量指标，得分与人类判断相关。代码可在该https URL获取。
英文摘要
Dense correspondence matching has historically been bounded by simplifying spatio-temporal priors, such as smooth motion and rigid geometry. While effective for classical tasks, these assumptions break down in image editing and reference-guided generation (IEG), where transformations can preserve visual identity while breaking physical continuity. To establish identity-preserving correspondence across such transformations, we introduce FreeMatching, a generalizable framework combining generative and semantic foundation representations with heterogeneous supervision from classical datasets, tracked videos, and synthetic scenes. Teacher-guided iterative refinement further improves correspondence in IEG without dense correspondence annotations. Experimentally, a single FreeMatching model substantially improves correspondence quality on challenging IEG image pairs while retaining competitive performance on classical benchmarks. Furthermore, we demonstrate its utility as a quantitative metric for evaluating identity preservation, with scores that correlate with human judgment. The code is available at https://github.com/luping-liu/FreeMatching.
CommentsAccepted at NeurIPS 2026. 24 pages, 7 figures, including appendices
