Research map / 3D vision / 3D generation
OVSeg3R: Learn Open-vocabulary Instance Segmentation from 2D via 3D Reconstruction
Hongyang Li, Jinyuan Qu, Lei 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 |
|---|---|---|---|
| add View-wise Instance Partition (ablation) | OVSeg3R · ScanNet200 standard setting | mAP: 18.4 -> 23.9 | Improved |
| add View-wise Instance Partition (ablation) | OVSeg3R · ScanNet200 standard setting | mAP novel: 16.2 -> 23.0 | Improved |
| use image-based superpoints (IBSp) instead of geometric-only superpoints (ablation) | OVSeg3R · ScanNet200 standard setting | mAP: 23.6 -> 23.9 | Improved |
| increase reconstructed data volume from 0% to 100% (ablation) | OVSeg3R · ScanNet200 standard setting | mAP: 5.0 -> 23.9 | Improved |
| increase reconstructed data volume from 0% to 100% (ablation) | OVSeg3R · ScanNet200 standard setting | mAP novel: 4e-4 -> 23.0 | Improved |
| add ScanNet3R-VGGT data alongside MASt3R-SLAM data (ablation) | OVSeg3R · ScanNet200 standard setting | mAP: 23.9 -> 24.6 | Improved |
| filter out failed reconstructions (ablation) | OVSeg3R · ScanNet200 standard setting | mAP: 23.0 -> 23.9 | Improved |
| use DINO-X universal mode features instead of 20-class prompts (ablation) | OVSeg3R · ScanNet200 standard setting | mAP novel: 14.0 -> 23.0 | Improved |
| use DINO-X universal mode features instead of 20-class prompts (ablation) | OVSeg3R · ScanNet200 standard setting | mAP base: 23.9 -> 26.2 | Improved |
| train SegDINO3D-VL with OVSeg3R instead of ScanNet200 only | SegDINO3D-VL · ScanNet200 val (open setting) | mAP: 38.4 -> 40.7 | Improved |
| train SegDINO3D-VL with OVSeg3R instead of ScanNet200 only | SegDINO3D-VL · ScanNet200 val (Tail classes) | mAP: 34.0 -> 42.7 | Improved |
| replace classification head with text-similarity head, trained on ScanNet200 only | SegDINO3D · ScanNet200 val | mAP: 39.8 -> 38.4 | Hurt |
Closest papers by meaning
- Open3DIS: Open-Vocabulary 3D Instance Segmentation with 2D Mask Guidance (CVPR 2024 · similarity 0.91)
- Segment Any 3D Object with Language (ICLR 2025 · similarity 0.90)
- Mosaic3D: Foundation Dataset and Model for Open-Vocabulary 3D Segmentation (CVPR 2025 · similarity 0.90)
- Open-YOLO 3D: Towards Fast and Accurate Open-Vocabulary 3D Instance Segmentation (ICLR 2025 · similarity 0.89)
- Open-Vocabulary 3D Semantic Segmentation with Foundation Models (CVPR 2024 · similarity 0.89)
- ImOV3D: Learning Open Vocabulary Point Clouds 3D Object Detection from Only 2D Images (NeurIPS 2024 · similarity 0.88)
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.
