Research map / 3D vision / Depth and stereo
SuperGlue: Learning Feature Matching With Graph Neural Networks
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 |
|---|---|---|---|
| Add SuperGlue matching to SuperPoint features instead of nearest-neighbor matching with OANet | SuperPoint · Indoor pose estimation (ScanNet) | Pose AUC @20°: 43.85 -> 51.84 | Helped |
| Add SuperGlue matching to SuperPoint features instead of nearest-neighbor matching with OANet | SuperPoint · Outdoor pose estimation (PhotoTourism) | Pose AUC @20°: 46.88 -> 64.16 | Helped |
| Replace nearest-neighbor plus OANet with SuperGlue | SuperPoint features · Homography estimation | AUC (DLT): 52.29 -> 65.85 | Helped |
| Add a graph neural network rather than use a single linear projection (ablation) | SuperGlue · Indoor pose estimation (ScanNet) | Pose AUC @20°: 38.56 -> 51.84 | Helped |
| Add cross-attention rather than replace it with self-attention (ablation) | SuperGlue · Indoor pose estimation (ScanNet) | Pose AUC @20°: 42.57 -> 51.84 | Helped |
| Add positional encoding with the keypoint encoder (ablation) | SuperGlue · Indoor pose estimation (ScanNet) | Pose AUC @20°: 47.12 -> 51.84 | Helped |
| Use 9 layers rather than 3 (ablation) | SuperGlue · Indoor pose estimation (ScanNet) | precision: 79.9 -> 84.4 | Helped |
| Backpropagate through the SuperPoint descriptor network during training (ablation) | SuperGlue with SuperPoint · Indoor pose estimation (ScanNet) | Pose AUC @20°: 51.84 -> 53.38 | Helped |
| Use SuperGlue instead of NN + mutual | SuperPoint features · Indoor pose estimation (ScanNet) | Pose AUC @20°: 36.40 -> 51.84 | Helped |
Closest papers by meaning
- DGC-GNN: Leveraging Geometry and Color Cues for Visual Descriptor-Free 2D-3D Matching (CVPR 2024 · similarity 0.87)
- SuperPrimitive: Scene Reconstruction at a Primitive Level (CVPR 2024 · similarity 0.86)
- FC-GNN: Recovering Reliable and Accurate Correspondences from Interferences (CVPR 2024 · similarity 0.85)
- OmniGlue: Generalizable Feature Matching with Foundation Model Guidance (CVPR 2024 · similarity 0.85)
- Matching While Perceiving: Enhance Image Feature Matching with Applicable Semantic Amalgamation (AAAI 2025 · similarity 0.84)
- SG-PGM: Partial Graph Matching Network with Semantic Geometric Fusion for 3D Scene Graph Alignment and Its Downstream Tasks (CVPR 2024 · similarity 0.84)
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.