Dense Match Summarization for Faster Two-view Estimation

TL;DR

This paper introduces a match summarization method that significantly accelerates robust two-view pose estimation using dense correspondences, maintaining accuracy while reducing runtime by up to 100 times.

Contribution

The paper presents a novel match summarization scheme that enables faster robust estimation without sacrificing accuracy in two-view geometry tasks.

Findings

Achieves 10-100x faster runtime compared to full dense match sets.

Maintains comparable accuracy to using all dense matches.

Validated on standard benchmarks with multiple dense matchers.

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.