Research map / 3D vision / 3D generation
Segment3D: Learning Fine-Grained Class-Agnostic 3D Segmentation without Manual Labels
Rui Huang, Songyou Peng, Ayca Takmaz, Federico Tombari, Marc Pollefeys, Shiji Song, Gao Huang, Francis Engelmann
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 |
|---|---|---|---|
| add self-supervised fine-tuning on full point clouds with high-confidence predicted masks (ablation) | Segment3D · ScanNet++ val | AP: 7.4 -> 12.0 | Improved |
| train on SAM-generated masks instead of manual 3D labels | Mask3D architecture · ScanNet++ val, small masks | AP50: 4.5 -> 15.9 | Improved |
| train on SAM-generated masks instead of manual 3D labels | Mask3D architecture · ScanNet++ val, large masks | AP50: 30.6 -> 30.4 | No clear effect |
| use a native 3D model on SAM labels instead of SAM3D merging | Segment3D · ScanNet++ val | AP: 3.9 -> 12.0 | Improved |
| increase number of queries from 100 to 400 (ablation) | Segment3D · ScanNet++ val | AP: 8.5 -> 12.0 | Improved |
| add 34k ScanNet++ frames to pre-training data (ablation) | Segment3D · ScanNet++ val | AP50: 22.7 -> 24.7 | Improved |
| replace Mask3D with Segment3D as mask generator in OpenMask3D | OpenMask3D · ScanNet++ open-set | AP50: 15.0 -> 17.7 | Improved |
| replace Mask3D with Segment3D, transfer indoor-trained models to outdoor scenes | Segment3D · Paris-Lille-3D | AP50: 1.5 -> 4.0 | Improved |
| vary number of queries for Mask3D (ablation) | Mask3D · ScanNet++ val | AP | No clear effect |
Later papers that compared with this method
1 later accepted paper reports Segment3D as a baseline in their results tables; 0 of them report a better number on most of the results they share with it.
- LR-AdaInSeg:Adaptive Instance Segmentation of Incomplete 3D Scenes Driven by Low-Rank Networks (AAAI 2026 · better on 0 of 2 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Any3DIS: Class-Agnostic 3D Instance Segmentation by 2D Mask Tracking (CVPR 2025 · similarity 0.88)
- Mosaic3D: Foundation Dataset and Model for Open-Vocabulary 3D Segmentation (CVPR 2025 · similarity 0.88)
- SAI3D: Segment Any Instance in 3D Scenes (CVPR 2024 · similarity 0.88)
- OnlineAnySeg: Online Zero-Shot 3D Segmentation by Visual Foundation Model Guided 2D Mask Merging (CVPR 2025 · similarity 0.88)
- Segment Any 3D Object with Language (ICLR 2025 · similarity 0.88)
- AGILE3D: Attention Guided Interactive Multi-object 3D Segmentation (ICLR 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.
