Research map / Detection and segmentation / Open-set and few-shot detection
SAS: Segment Any 3D Scene with Integrated 2D Priors
Zhuoyuan Li, Jiahao Lu, Jiacheng Deng, Hanzhi Chang, Lifan Wu, Yanzhe Liang, Tianzhu Zhang
Read the paperCheck 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 capability-guided fusion instead of simple addition of 2D features (ablation) | SAS · ScanNet v2 | mIoU: 48.6 -> 55.5 | Improved |
| use capability-guided fusion instead of linear fusion (ablation) | SAS · Matterport3D | mIoU: 39.4 -> 43.6 | Improved |
| use capability-guided fusion instead of linear fusion (ablation) | SAS · nuScenes | mIoU: 34.8 -> 40.0 | Improved |
| fuse LSeg and SEEM features rather than using LSeg alone (ablation) | SAS 2D features · ScanNet v2 | mIoU: 51.2 -> 55.5 | Improved |
| add superpoint distillation to pixel-point distillation (ablation) | SAS · ScanNet v2 | mIoU: 56.7 -> 59.2 | Improved |
| add superpoint distillation to pixel-point distillation (ablation) | SAS · nuScenes | mIoU: 45.4 -> 45.4 | No clear effect |
| add temporal ensembling self-distillation (ablation) | SAS · ScanNet v2 | mIoU: 59.2 -> 61.9 | Improved |
| add temporal ensembling self-distillation (ablation) | SAS · Matterport3D | mIoU: 46.3 -> 48.6 | Improved |
| add temporal ensembling self-distillation (ablation) | SAS · nuScenes | mIoU: 45.4 -> 47.5 | Improved |
| distill the fused 2D features into 3D model instead of using 2D features directly (pixel-point) (ablation) | SAS · nuScenes | mIoU: +5.4% | Improved |
| add ODISE as third 2D model | SAS · ScanNet v2 | mIoU: 61.9 -> 62.5 | Improved |
| add ODISE as third 2D model | SAS · Matterport3D | mIoU: 48.6 -> 49.8 | Improved |
Closest papers by meaning
- SAI3D: Segment Any Instance in 3D Scenes (CVPR 2024 · similarity 0.89)
- Mosaic3D: Foundation Dataset and Model for Open-Vocabulary 3D Segmentation (CVPR 2025 · similarity 0.88)
- Open3DIS: Open-Vocabulary 3D Instance Segmentation with 2D Mask Guidance (CVPR 2024 · similarity 0.87)
- Segment Any 3D Object with Language (ICLR 2025 · similarity 0.87)
- Cross-Modal and Uncertainty-Aware Agglomeration for Open-Vocabulary 3D Scene Understanding (CVPR 2025 · similarity 0.87)
- SA3DIP: Segment Any 3D Instance with Potential 3D Priors (NeurIPS 2024 · similarity 0.87)
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.
