Research map / Detection and segmentation / CNN features
Reinforced Feature Points: Optimizing Feature Detection and Description for a High-Level Task
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Train SuperPoint end-to-end using relative pose error as the task loss | SuperPoint · Relative pose estimation | pose accuracy | Helped |
| Train SuperPoint end-to-end using relative pose error as the task loss | SuperPoint · H-Patches | matching accuracy | Helped |
| Train SuperPoint for relative pose estimation | SuperPoint · Fountain | reprojection error (px): 0.97 -> 0.87 | Helped |
| Train SuperPoint for relative pose estimation | SuperPoint · Herzjesu | reprojection error (px): 0.95 -> 0.82 | Helped |
| Train SuperPoint for relative pose estimation | SuperPoint · South Building | reprojection error (px): 0.92 -> 0.88 | Helped |
| Train SuperPoint for relative pose estimation | SuperPoint · Outdoor relative pose estimation | inlier ratio: 24.8% -> 28.4% | Helped |
| Train SuperPoint for relative pose estimation | SuperPoint · Indoor relative pose estimation | inlier ratio: 13.4% -> 16.4% | Helped |
| Jointly train SuperPoint and NG-RANSAC rather than training only one (ablation) | SuperPoint with NG-RANSAC · Relative pose estimation | pose accuracy | Helped |
| Train for relative pose estimation rather than only low-level matching scores | SuperPoint · Relative pose estimation | pose accuracy | Mixed |
Closest papers by meaning
- Collaborative Feature Matching with Progressive Correspondence Learning (AAAI 2026 · similarity 0.89)
- XFeat: Accelerated Features for Lightweight Image Matching (CVPR 2024 · similarity 0.88)
- OmniGlue: Generalizable Feature Matching with Foundation Model Guidance (CVPR 2024 · similarity 0.87)
- Towards Complementary Keypoint Detection via Mixture of Detectors (NeurIPS 2026 · similarity 0.87)
- MATCHA: Towards Matching Anything (CVPR 2025 · similarity 0.87)
- LookThere! Sparse Vision by Reinforced Selection (NeurIPS 2026 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.