Research map / Point clouds / Registration and matching
Self-Supervised Geometric Perception
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 CAPS using SGP pseudo-labels rather than ground-truth labels | CAPS · MegaDepth test | rotation and translation recall | Helped |
| Train FCGF using SGP pseudo-labels rather than ground-truth labels | FCGF · 3DMatch test | registration recall: 78.44 -> 91.37 | Helped |
| Train FCGF using SGP pseudo-labels rather than ground-truth labels | FCGF · 3DMatch test | registration recall | Helped |
| Replace finetuning with retraining from random weights at each iteration (ablation) | SGP with FCGF · 3DMatch test | registration recall: 91.2% and 90.9% | No clear effect |
| Remove the verifier (ablation) | SGP with FCGF · 3DMatch test | registration recall: 91.4% and 90.6% | No clear effect |
| Train SGP on the smaller test set instead of the larger training set (ablation) | SGP with FCGF · 3DMatch test set and training set | registration recall: 91.37%; below 82%; over 92% | Mixed |
| Replace robust RANSAC10K teacher with non-robust Horn's method (ablation) | SGP with FCGF · 3DMatch | test registration recall: below 30% | Hurt |
Closest papers by meaning
- Improving Semantic Correspondence with Viewpoint-Guided Spherical Maps (CVPR 2024 · similarity 0.86)
- Less Decoder is More Encoder: Geometric Representation Learning from Novel View Synthesis (NeurIPS 2026 · similarity 0.85)
- DualNet: Robust Self-Supervised Stereo Matching with Pseudo-Label Supervision (AAAI 2025 · similarity 0.85)
- Self-Supervised Cross-View Correspondence with Predictive Cycle Consistency (CVPR 2025 · similarity 0.85)
- Probing the Mid-level Vision Capabilities of Self-Supervised Learning (CVPR 2025 · similarity 0.85)
- Mining and Transferring Feature-Geometry Coherence for Unsupervised Point Cloud Registration (NeurIPS 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.