Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper addresses limitations in semi-dense image matching, specifically the over-exclusion issue in mutual nearest neighbor (MNN) matching when scale differences exist and the lack of local consistency in fine-stage matching. They introduce a scale-aware matching module that leverages information within the score matrix to estimate scale ratios, improving coarse matching. Furthermore, they reformulate the fine stage as a cascaded flow refinement problem and introduce a gradient loss to encourage local consistency.
Scale differences no longer doom image matching: a novel scale-aware matching module exploits score matrix hints to drastically improve performance with negligible overhead.
Recent semi-dense image matching methods have achieved remarkable success, but two long-standing issues still impair their performance. At the coarse stage, the over-exclusion issue of their mutual nearest neighbor (MNN) matching layer makes them struggle to handle cases with scale difference between images. To this end, we comprehensively revisit the matching mechanism and make a key observation that the hint concealed in the score matrix can be exploited to indicate the scale ratio. Based on this, we propose a scale-aware matching module which is exceptionally effective but introduces negligible overhead. At the fine stage, we point out that existing methods neglect the local consistency of final matches, which undermines their robustness. To this end, rather than independently predicting the correspondence for each source pixel, we reformulate the fine stage as a cascaded flow refinement problem and introduce a novel gradient loss to encourage local consistency of the flow field. Extensive experiments demonstrate that our novel matching pipeline, with these proposed modifications, achieves robust and accurate matching performance on downstream tasks.