الملخص
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.
الكلمات المفتاحية
الموضوع
بيانات النشر
- المجلة
- غير متاح
- وصول مفتوح
- وصول مفتوح أخضر
اقتبس هذه المقالة
APA 7
Liu, L., Kang, B., Wang, Y., & Xu, D. (2026). Beyond Spatio-Temporal Priors: A Generalizable Approach for Dense Correspondence Matching. https://omanscience.com/ar/articles/beyond-spatio-temporal-priors-a-generalizable-approach-for-dense-correspondence-matching
MLA 9
Liu, Luping, et al. "Beyond Spatio-Temporal Priors: A Generalizable Approach for Dense Correspondence Matching." https://omanscience.com/ar/articles/beyond-spatio-temporal-priors-a-generalizable-approach-for-dense-correspondence-matching.
شيكاغو (المؤلف–التاريخ)
Liu, Luping, Bingyi Kang, Yifan Wang, and Dong Xu. 2026. "Beyond Spatio-Temporal Priors: A Generalizable Approach for Dense Correspondence Matching." https://omanscience.com/ar/articles/beyond-spatio-temporal-priors-a-generalizable-approach-for-dense-correspondence-matching.
هارفارد
Liu, L., Kang, B., Wang, Y. and Xu, D. (2026) 'Beyond Spatio-Temporal Priors: A Generalizable Approach for Dense Correspondence Matching', Available at: https://omanscience.com/ar/articles/beyond-spatio-temporal-priors-a-generalizable-approach-for-dense-correspondence-matching.
فانكوفر
Liu L, Kang B, Wang Y, Xu D. Beyond Spatio-Temporal Priors: A Generalizable Approach for Dense Correspondence Matching. https://omanscience.com/ar/articles/beyond-spatio-temporal-priors-a-generalizable-approach-for-dense-correspondence-matching
IEEE
L. Liu, B. Kang, Y. Wang, and D. Xu, "Beyond Spatio-Temporal Priors: A Generalizable Approach for Dense Correspondence Matching," https://omanscience.com/ar/articles/beyond-spatio-temporal-priors-a-generalizable-approach-for-dense-correspondence-matching.