Dense Match Summarization for Faster Two-View Estimation

Jonathan Astermark, Anders Heyden, Viktor Larsson

CVPR 2025 pp. 1093-1102

doi:10.1109/CVPR52734.2025.00110 /cvpr/2025/astermark2025cvpr-dense/

Abstract

In this paper, we speed up robust two-view relative pose from dense correspondences. Previous work has shown that dense matchers can significantly improve both accuracy and robustness in the resulting pose. However, the large number of matches comes with a significantly increased runtime during robust estimation in RANSAC. To avoid this, we propose an efficient match summarization scheme which provides comparable accuracy to using the full set of dense matches, while having 10-100x faster runtime. We validate our approach on standard benchmark datasets together with multiple state-of-the-art dense matchers.

PDF CVPR Semantic Scholar

Cite

Text

Astermark et al. "Dense Match Summarization for Faster Two-View Estimation." Conference on Computer Vision and Pattern Recognition, 2025. doi:10.1109/CVPR52734.2025.00110

Markdown

[Astermark et al. "Dense Match Summarization for Faster Two-View Estimation." Conference on Computer Vision and Pattern Recognition, 2025.](https://mlanthology.org/cvpr/2025/astermark2025cvpr-dense/) doi:10.1109/CVPR52734.2025.00110

BibTeX

@inproceedings{astermark2025cvpr-dense,
  title     = {{Dense Match Summarization for Faster Two-View Estimation}},
  author    = {Astermark, Jonathan and Heyden, Anders and Larsson, Viktor},
  booktitle = {Conference on Computer Vision and Pattern Recognition},
  year      = {2025},
  pages     = {1093-1102},
  doi       = {10.1109/CVPR52734.2025.00110},
  url       = {https://mlanthology.org/cvpr/2025/astermark2025cvpr-dense/}
}