Research map / Label-efficient and robust learning / Self-supervised and contrastive
CONSAC: Robust Multi-Model Fitting by Conditional Sample Consensus
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 |
|---|---|---|---|
| Use CONSAC rather than T-Linkage | vanishing point estimation · NYU-VP | AUC (%): 57.8 -> 65.0 | Helped |
| Use CONSAC rather than T-Linkage | vanishing point estimation · YUD+ | AUC (%): 72.6 -> 77.1 | Helped |
| Use CONSAC rather than Kluger | vanishing point estimation · YUD | AUC (%): 85.9 -> 83.9 | Hurt |
| Use CONSAC-S rather than Progressive-X | multi-homography estimation · AdelaideRMF-H | ME (%): 6.86 -> 5.21 | Helped |
| Add EM refinement (ablation) | CONSAC · NYU-VP | AUC: 62.90 -> 65.01 | Helped |
| Add EM refinement (ablation) | CONSAC-S · AdelaideRMF | ME: 6.17 -> 5.21 | Helped |
| Use supervised rather than self-supervised training (ablation) | CONSAC · NYU-VP | AUC: 63.44 -> 65.01 | Helped |
| Add inlier masking regularisation (ablation) | CONSAC-S · NYU-VP | AUC: 59.94 -> 63.44 | Helped |
| Add inlier masking regularisation (ablation) | CONSAC-S · AdelaideRMF | ME: 8.14 -> 5.21 | Helped |
| Use the self-supervised loss in addition to inlier masking regularisation (ablation) | CONSAC-S · NYU-VP | AUC: 29.31 -> 63.44 | Helped |
| Use the self-supervised loss in addition to inlier masking regularisation (ablation) | CONSAC-S · AdelaideRMF | ME: 21.12 -> 5.21 | Helped |
| Use conditional rather than unconditional sampling (ablation) | CONSAC · NYU-VP | AUC: 48.36 -> 65.01 | Helped |
Closest papers by meaning
- PARSAC: Accelerating Robust Multi-Model Fitting with Parallel Sample Consensus (AAAI 2024 · similarity 0.92)
- On the Instability of Relative Pose Estimation and RANSAC's Role (CVPR 2022 · similarity 0.88)
- Latent RANSAC (CVPR 2018 · similarity 0.88)
- Noisy One-point Homographies are Surprisingly Good (CVPR 2024 · similarity 0.87)
- Convex Relaxation for Robust Vanishing Point Estimation in Manhattan World (CVPR 2025 · similarity 0.86)
- RESfM: Robust Deep Equivariant Structure from Motion (ICLR 2025 · similarity 0.83)
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.