Research map / Detection and segmentation / Open-set and few-shot detection
Multimodal Material Segmentation
Yupeng Liang, Ryosuke Wakaki, Shohei Nobuhara, Ko Nishino
Check your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 MCubeSNet with region-guided filter selection on multimodal input | Material segmentation on MCubeS · MCubeS test | mIoU: 40.58% -> 42.86% | Improved |
| Add class weight balancing to network outputs | MCubeSNet · MCubeS test | mIoU: 42.86% -> 45.95% | Improved |
| Set RGFSConv ratio λ to 1 instead of 3 (ablation) | MCubeSNet · MCubeS test | mIoU: 42.86% -> 38.13% | Hurt |
| Increase RGFSConv ratio λ above 3 (ablation) | MCubeSNet · MCubeS test | mIoU: 42.86% -> 39.54%; 42.86% -> 39.13% | Hurt |
| Place RGFSConv at the second rather than first decoder layer (ablation) | MCubeSNet · MCubeS test | mIoU: 42.86% -> 39.65% | Hurt |
| Use RGFSConv at both decoder layers (ablation) | MCubeSNet · MCubeS test | mIoU: 42.86% -> 33.08% | Hurt |
| Add NIR to RGB, AoLP, and DoLP (ablation) | MCubeSNet · MCubeS test | mIoU: 39.5 -> 42.9 | Improved |
| Add AoLP and DoLP to RGB (ablation) | MCubeSNet · MCubeS test | mIoU: 33.7 -> 39.5 | Improved |
| Add AoLP to RGB (ablation) | MCubeSNet · MCubeS test | mIoU: 33.7 -> 39.1 | Improved |
| Add NIR to RGB (ablation) | MCubeSNet · MCubeS test | mIoU: 33.7 -> 40.5 | Improved |
| Use RGB, DoLP, and NIR rather than RGB and DoLP (ablation) | MCubeSNet · MCubeS test | mIoU: 34.1 -> 42.0 | Improved |
Closest papers by meaning
- Infusing Synthetic Data with Real-World Patterns for Zero-Shot Material State Segmentation (NeurIPS 2024 · similarity 0.85)
- Keep the Balance: A Parameter-Efficient Symmetrical Framework for RGB+X Semantic Segmentation (CVPR 2025 · similarity 0.84)
- Exploiting Polarized Material Cues for Robust Car Detection (AAAI 2024 · similarity 0.83)
- UNIP: Rethinking Pre-trained Attention Patterns for Infrared Semantic Segmentation (ICLR 2025 · similarity 0.83)
- Material Palette: Extraction of Materials from a Single Image (CVPR 2024 · similarity 0.83)
- PARTONOMY: Large Multimodal Models with Part-Level Visual Understanding (NeurIPS 2025 · 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.
