UFM: A Simple Path towards Unified Dense Correspondence with Flow

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Zhang, Yuchen, Keetha, Nikhil, Lyu, Chenwei, Jhamb, Bhuvan, Chen, Yutian, Qiu, Yuheng, Karhade, Jay, Jha, Shreyas, Hu, Yaoyu, Ramanan, Deva, Scherer, Sebastian, Wang, Wenshan
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866908823926603776
author Zhang, Yuchen
Keetha, Nikhil
Lyu, Chenwei
Jhamb, Bhuvan
Chen, Yutian
Qiu, Yuheng
Karhade, Jay
Jha, Shreyas
Hu, Yaoyu
Ramanan, Deva
Scherer, Sebastian
Wang, Wenshan
author_facet Zhang, Yuchen
Keetha, Nikhil
Lyu, Chenwei
Jhamb, Bhuvan
Chen, Yutian
Qiu, Yuheng
Karhade, Jay
Jha, Shreyas
Hu, Yaoyu
Ramanan, Deva
Scherer, Sebastian
Wang, Wenshan
contents Dense image correspondence is central to many applications, such as visual odometry, 3D reconstruction, object association, and re-identification. Historically, dense correspondence has been tackled separately for wide-baseline scenarios and optical flow estimation, despite the common goal of matching content between two images. In this paper, we develop a Unified Flow & Matching model (UFM), which is trained on unified data for pixels that are co-visible in both source and target images. UFM uses a simple, generic transformer architecture that directly regresses the (u,v) flow. It is easier to train and more accurate for large flows compared to the typical coarse-to-fine cost volumes in prior work. UFM is 28% more accurate than state-of-the-art flow methods (Unimatch), while also having 62% less error and 6.7x faster than dense wide-baseline matchers (RoMa). UFM is the first to demonstrate that unified training can outperform specialized approaches across both domains. This result enables fast, general-purpose correspondence and opens new directions for multi-modal, long-range, and real-time correspondence tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2506_09278
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle UFM: A Simple Path towards Unified Dense Correspondence with Flow
Zhang, Yuchen
Keetha, Nikhil
Lyu, Chenwei
Jhamb, Bhuvan
Chen, Yutian
Qiu, Yuheng
Karhade, Jay
Jha, Shreyas
Hu, Yaoyu
Ramanan, Deva
Scherer, Sebastian
Wang, Wenshan
Computer Vision and Pattern Recognition
Machine Learning
Robotics
Dense image correspondence is central to many applications, such as visual odometry, 3D reconstruction, object association, and re-identification. Historically, dense correspondence has been tackled separately for wide-baseline scenarios and optical flow estimation, despite the common goal of matching content between two images. In this paper, we develop a Unified Flow & Matching model (UFM), which is trained on unified data for pixels that are co-visible in both source and target images. UFM uses a simple, generic transformer architecture that directly regresses the (u,v) flow. It is easier to train and more accurate for large flows compared to the typical coarse-to-fine cost volumes in prior work. UFM is 28% more accurate than state-of-the-art flow methods (Unimatch), while also having 62% less error and 6.7x faster than dense wide-baseline matchers (RoMa). UFM is the first to demonstrate that unified training can outperform specialized approaches across both domains. This result enables fast, general-purpose correspondence and opens new directions for multi-modal, long-range, and real-time correspondence tasks.
title UFM: A Simple Path towards Unified Dense Correspondence with Flow
topic Computer Vision and Pattern Recognition
Machine Learning
Robotics
url https://arxiv.org/abs/2506.09278