Research map / 3D vision / Face and shape
Geometry-Informed Material Recognition
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 joint texture-normal features to FV+CNN (ablation) | FV+CNN · GeoMat focus-scale test | mean accuracy: 68.92 -> 73.80 | Helped |
| Add independent normal features to FV-N+CNN (ablation) | FV-N+CNN · GeoMat focus-scale test | mean accuracy: 73.80 -> 73.84 | Helped |
| Add independent normal features to FV+CNN (ablation) | FV+CNN · GeoMat focus-scale test | mean accuracy: 68.92 -> 72.08 | Helped |
| Add HSV features to FV+CNN (ablation) | FV+CNN · GeoMat focus-scale test | mean accuracy: 68.92 -> 67.82 | Hurt |
| Add HSV features to FV-N+CNN (ablation) | FV-N+CNN · GeoMat focus-scale test | mean accuracy: 73.80 -> 72.05 | Hurt |
| Jointly model geometry with texture (ablation) | RFS, MR8, and FV feature sets · GeoMat focus-scale test | mean accuracy: 33.24 -> 37.76; 32.47 -> 41.34; 60.97 -> 66.95 | Helped |
| Add independent normal histograms (ablation) | RFS, MR8, and FV feature sets · GeoMat focus-scale test | mean accuracy: 33.24 -> 49.68; 32.47 -> 49.74; 60.97 -> 65.87 | Helped |
| Add HSV features (ablation) | RFS, MR8, and FV feature sets · GeoMat focus-scale test | mean accuracy: 33.24 -> 45.03; 32.47 -> 45.32; 60.97 -> 62.92 | Helped |
| Frontally rectify patches based on mean surface normal (ablation) | FV-N and FV-N+CNN · GeoMat focus-scale test | mean accuracy: 66.95 -> 66.82; 73.80 -> 72.97 | Hurt |
| Frontally rectify patches based on mean surface normal (ablation) | RFS and MR8 filter features · GeoMat focus-scale test | mean accuracy: 33.24 -> 34.42; 32.47 -> 35.03 | Helped |
| Use 3D features instead of the best 2D feature set | FV+CNN · GeoMat scene-scale test | pixel labeling accuracy: 21.01 -> 35.87 | Helped |
| Use 3D features instead of the best 2D feature set | FV+CNN · GeoMat focus-scale test | mean accuracy: 68.92 -> 73.84 | Helped |
Closest papers by meaning
- Infusing Synthetic Data with Real-World Patterns for Zero-Shot Material State Segmentation (NeurIPS 2024 · similarity 0.83)
- 3DCoMPaT200: Language Grounded Large-Scale 3D Vision Dataset for Compositional Recognition (NeurIPS 2024 · similarity 0.82)
- PolyhedronNet: Representation Learning for Polyhedra with Surface-attributed Graph (ICLR 2025 · similarity 0.82)
- GeoX: Geometric Problem Solving Through Unified Formalized Vision-Language Pre-training (ICLR 2025 · similarity 0.82)
- Synthesizing Multimodal Geometry Datasets from Scratch and Enabling Visual Alignment via Plotting Code (ICML 2026 · similarity 0.82)
- Cracks in the Foundation: A Civil Infrastructure Dataset to Challenge Vision Foundation Models (NeurIPS 2026 · similarity 0.82)
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.