Research map / 3D vision / Gaussian splatting
LEGO: Leveled Language Gaussian Splatting
Yuning Peng, Haiping Wang, Yuan Liu, Yipeng Lu, Zhen Dong, Bisheng Yang
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 scale-balancing re-weighting to the contrastive loss (ablation) | LEGO · Mip-NeRF 360 Room | mAcc / mIoU: 86.2 -> 93.1 mAcc; 63.6 -> 67.3 mIoU | Improved |
| add positive L2 penalty L_pos (ablation) | LEGO · Mip-NeRF 360 Room | mIoU: 62.3 -> 63.6 | Improved |
| add feature normalization L_3d/L_2d (ablation) | LEGO · Mip-NeRF 360 Room | mIoU: 60.7 -> 62.3 | Improved |
| add LLM-guided chain-of-retrieval over the level-wise scene graph (ablation) | LEGO · CoR queries | mIoU: 22.7 -> 51.6 | Improved |
| add the same CoR strategy to LaGa | LaGa · CoR queries | mIoU: 8.3 -> 14.3 | Improved |
| use 3D level-based hierarchy instead of SAM 3-level granularities | LEGO vs LaGa w/ CoR · CoR queries | mIoU: 14.3 -> 51.6 | Improved |
| level-wise decoupled features vs prior open-vocabulary fields | LEGO · LERF-OVS | 3D segmentation mIoU: 64.0 -> 68.4 | Improved |
| level-wise features vs prior methods on promptable segmentation | LEGO · SPIn-NeRF | mIoU: 94.3 -> 94.2 | No clear effect |
| level-wise features vs SAGA on NVOS | LEGO · NVOS | mIoU: 92.6 -> 94.2 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 5 on most shared results.
- SAGA — SAGA: Stochastic Whole-Body Grasping with Contact (ECCV 2022) · better on 3 of 3 shared results
- LangSplatV2 — LangSplatV2: High-dimensional 3D Language Gaussian Splatting with 450+ FPS (NeurIPS 2025) · better on 2 of 2 shared results
- GAGS — GAGS: Granularity-Aware Feature Distillation for Language Gaussian Splatting (AAAI 2026) · better on 1 of 1 shared result
- LangSplat — LangSplat: 3D Language Gaussian Splatting (CVPR 2024) · better on 1 of 1 shared result
- OmniSeg3D — OmniSeg3D: Omniversal 3D Segmentation via Hierarchical Contrastive Learning (CVPR 2024) · better on 0 of 1 shared result
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
- Segment then Splat: Unified 3D Open-Vocabulary Segmentation via Gaussian Splatting (NeurIPS 2025 · similarity 0.89)
- PanoGS: Gaussian-based Panoptic Segmentation for 3D Open Vocabulary Scene Understanding (CVPR 2025 · similarity 0.89)
- OpenLex3D: A Tiered Benchmark for Open-Vocabulary 3D Scene Representations (NeurIPS 2025 · similarity 0.88)
- Mosaic3D: Foundation Dataset and Model for Open-Vocabulary 3D Segmentation (CVPR 2025 · similarity 0.87)
- LangSplat: 3D Language Gaussian Splatting (CVPR 2024 · similarity 0.87)
- 3D Vision-Language Gaussian Splatting (ICLR 2025 · 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.
