Research map / 3D vision / Gaussian splatting
FMGS: Foundation Model Embedded 3D Gaussian Splatting for Holistic 3D Scene Understanding
Xingxing Zuo, Pouya Samangouei, Yunwen Zhou, Yan Di, Mingyang Li
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 pixel alignment (dot-product similarity) loss (ablation) | FMGS · LERF object detection (5 scenes) | average accuracy (%): 90.4 -> 93.2 | Improved |
| add hybrid multi-scale CLIP supervision instead of single finest-scale CLIP (ablation) | FMGS · LERF object detection (5 scenes) | average accuracy (%): 30.8 -> 93.2 | Improved |
| use single hybrid CLIP supervision instead of LERF multi-scale CLIP (ablation) | FMGS · LERF object detection (5 scenes) | average accuracy (%): 78.8 -> 93.2 | Improved |
| add multi-resolution hash encoding to Gaussian scene representation (ablation) | FMGS · LERF object detection (5 scenes) | average accuracy (%): 84.4 -> 93.2 | Improved |
| embed CLIP and DINO features in 3D Gaussians with hash encoding instead of NeRF (LERF) | FMGS vs LERF · LERF object detection (5 scenes) | average accuracy (%): 83.0 -> 93.2 | Improved |
| replace NeRF rendering with Gaussian splatting for inference | FMGS vs LERF · LERF object detection (5 scenes) | inference FPS: 0.1214 -> 103.4 | Improved |
| use FMGS instead of LERF for unsupervised segmentation | FMGS vs LERF · 3D-OVS segmentation | mIoU: bed 33.5 -> 38.0; sofa 28.1 -> 56.6 | Improved |
| refine segmentation with SAM masks and majority voting | FMGS · 3D-OVS segmentation | mIoU: bed 38.0 -> 80.6; sofa 56.6 -> 90.8 | Improved |
| distill CLIP features into 3D Gaussians instead of LSeg features | FMGS vs FFD-LSeg · LERF object detection (5 scenes) | average accuracy (%): 18.0 -> 93.2 | Improved |
Later papers that compared with this method
1 later accepted paper reports FMGS as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- 3D Vision-Language Gaussian Splatting (ICLR 2025 · better on 1 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- LERF — LERF: Language Embedded Radiance Fields (ICCV 2023) · better on 13 of 14 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
- SceneSplat++: A Large Dataset and Comprehensive Benchmark for Language Gaussian Splatting (NeurIPS 2025 · similarity 0.91)
- FHGS: Feature-Homogenized Gaussian Splatting (NeurIPS 2025 · similarity 0.91)
- Dr. Splat: Directly Referring 3D Gaussian Splatting via Direct Language Embedding Registration (CVPR 2025 · similarity 0.90)
- Feature 3DGS: Supercharging 3D Gaussian Splatting to Enable Distilled Feature Fields (CVPR 2024 · similarity 0.89)
- Bootstraping Clustering of Gaussians for View-consistent 3D Scene Understanding (AAAI 2025 · similarity 0.89)
- SfM-Free 3D Gaussian Splatting via Hierarchical Training (CVPR 2025 · similarity 0.89)
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.
