Research map / 3D vision / Depth and stereo
Unsupervised Learning of Consensus Maximization for 3D Vision Problems
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 |
|---|---|---|---|
| Fine-tune synthetic-data training with unsupervised consensus loss | synth-sv · FAUST intra-subject | inliers: 3812 -> 3814 | Helped |
| Fine-tune synthetic-data training with unsupervised consensus loss | synth-sv · FAUST intra-subject | outliers: 74 -> 58 | Helped |
| Fine-tune synthetic-data training with unsupervised consensus loss | synth-sv · FAUST inter-subject | inliers: 2601 -> 3122 | Helped |
| Fine-tune synthetic-data training with unsupervised consensus loss | synth-sv · FAUST inter-subject | outliers: 82 -> 11 | Helped |
| Fine-tune synthetic-data training with unsupervised consensus loss | synth-sv · FAUST inter-subject | inliers: 2371 -> 3303 | Helped |
| Fine-tune synthetic-data training with unsupervised consensus loss | synth-sv · FAUST inter-subject | outliers: 171 -> 40 | Helped |
| Use piecewise rigidity as the deformation model | Unsupervised consensus maximization · FAUST non-rigid shape matching | inlier matches | Mixed |
| Train on data-specific correspondence statistics rather than relying only on synthetic pretraining | Unsupervised consensus maximization · FAUST intra- and inter-subject | outlier rejection: 74 -> 58; 82 -> 11 | Helped |
| Train on synthetic data alone | Consensus maximization network · Real data with different distributions | classification accuracy | Hurt |
| Use a single Fundamental matrix model on homography data | 2D-2D geometry estimation · Homography correspondences | outlier rejection | Hurt |
| Increase RANSAC maximum iterations to handle higher outlier rates | USAC · ModelNet-40 homography estimation | runtime | Hurt |
Closest papers by meaning
- Guaranteed Outlier Removal With Mixed Integer Linear Programs (CVPR 2016 · similarity 0.88)
- Consensus Maximization With Linear Matrix Inequality Constraints (CVPR 2017 · similarity 0.87)
- Scalable 3D Registration via Truncated Entry-wise Absolute Residuals (CVPR 2024 · similarity 0.86)
- PARSAC: Accelerating Robust Multi-Model Fitting with Parallel Sample Consensus (AAAI 2024 · similarity 0.86)
- Convex Relaxation for Robust Vanishing Point Estimation in Manhattan World (CVPR 2025 · similarity 0.85)
- Unsupervised Learning of Category-Level 3D Pose from Object-Centric Videos (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.