Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2601.12530 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911383921098752 |
|---|---|
| author | Schmid, Jan Fabian Hagemann, Annika |
| author_facet | Schmid, Jan Fabian Hagemann, Annika |
| contents | Sparse keypoint matching is crucial for 3D vision tasks, yet current keypoint detectors often produce spatially inaccurate matches. Existing refinement methods mitigate this issue through alignment of matched keypoint locations, but they are typically detector-specific, requiring retraining for each keypoint detector. We introduce XRefine, a novel, detector-agnostic approach for sub-pixel keypoint refinement that operates solely on image patches centered at matched keypoints. Our cross-attention-based architecture learns to predict refined keypoint coordinates without relying on internal detector representations, enabling generalization across detectors. Furthermore, XRefine can be extended to handle multi-view feature tracks. Experiments on MegaDepth, KITTI, and ScanNet demonstrate that the approach consistently improves geometric estimation accuracy, achieving superior performance compared to existing refinement methods while maintaining runtime efficiency. Our code and trained models can be found at https://github.com/boschresearch/xrefine. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_12530 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | XRefine: Attention-Guided Keypoint Match Refinement Schmid, Jan Fabian Hagemann, Annika Computer Vision and Pattern Recognition Sparse keypoint matching is crucial for 3D vision tasks, yet current keypoint detectors often produce spatially inaccurate matches. Existing refinement methods mitigate this issue through alignment of matched keypoint locations, but they are typically detector-specific, requiring retraining for each keypoint detector. We introduce XRefine, a novel, detector-agnostic approach for sub-pixel keypoint refinement that operates solely on image patches centered at matched keypoints. Our cross-attention-based architecture learns to predict refined keypoint coordinates without relying on internal detector representations, enabling generalization across detectors. Furthermore, XRefine can be extended to handle multi-view feature tracks. Experiments on MegaDepth, KITTI, and ScanNet demonstrate that the approach consistently improves geometric estimation accuracy, achieving superior performance compared to existing refinement methods while maintaining runtime efficiency. Our code and trained models can be found at https://github.com/boschresearch/xrefine. |
| title | XRefine: Attention-Guided Keypoint Match Refinement |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2601.12530 |