Research map / 3D vision / Gaussian splatting
LightSplat: Fast and Memory-Efficient Open-Vocabulary 3D Scene Understanding in Five Seconds
Jaehun Bang, Jinhyeok Kim, Minji Kim, Seungheon Jeong, Kyungdon Joo
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 3D-aware mask filtering (ablation) | LightSplat · LERF-OVS | mIoU: 29.31 -> 47.58 | Improved |
| add 3D-aware mask filtering (ablation) | LightSplat · LERF-OVS | mAcc@0.25: 25.96 -> 68.32 | Improved |
| add semantic-aware clustering (ablation) | LightSplat · LERF-OVS | mIoU: 19.56 -> 47.58 | Improved |
| add semantic-aware clustering (ablation) | LightSplat · LERF-OVS | mAcc@0.25: 18.97 -> 68.32 | Improved |
| add geometry-aware clustering (ablation) | LightSplat · LERF-OVS | mIoU: 2.01 -> 47.58 | Improved |
| add geometry-aware clustering (ablation) | LightSplat · LERF-OVS | mAcc@0.25: 2.27 -> 68.32 | Improved |
| add the ablated components (filtering and clustering) to the pipeline (ablation) | LightSplat · LERF-OVS | feature distillation time: 4.42 s -> 4.55 s | No clear effect |
| replace OpenGaussian-style iterative per-Gaussian features with index-feature mapping and clustering | LightSplat · LERF-OVS | mIoU: 42.15 -> 47.58 | Improved |
| replace OpenGaussian-style iterative per-Gaussian features with index-feature mapping and clustering | LightSplat · ScanNet (19 classes) | mIoU: 29.43 -> 37.11 | Improved |
| replace LUDVIG per-Gaussian 2048-byte features with 2-byte indices | LightSplat · ScanNet | memory per Gaussian (bytes): 2048 -> 2 | Improved |
| replace LUDVIG iterative aggregation with single-step clustering | LightSplat · DL3DV-OVS | mIoU: 29.21 -> 44.98 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- LUDVIG — LUDVIG: Learning-Free Uplifting of 2D Visual Features to Gaussian Splatting Scenes (ICCV 2025) · better on 6 of 6 shared results
- LangSplat — LangSplat: 3D Language Gaussian Splatting (CVPR 2024) · better on 4 of 4 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
- FastLGS: Speeding Up Language Embedded Gaussians with Feature Grid Mapping (AAAI 2025 · similarity 0.90)
- OpenLex3D: A Tiered Benchmark for Open-Vocabulary 3D Scene Representations (NeurIPS 2025 · similarity 0.89)
- RegionPLC: Regional Point-Language Contrastive Learning for Open-World 3D Scene Understanding (CVPR 2024 · similarity 0.89)
- Rethinking Open-Vocabulary Segmentation of Radiance Fields in 3D Space (AAAI 2025 · similarity 0.89)
- Tackling View-Dependent Semantics in 3D Language Gaussian Splatting (ICML 2025 · similarity 0.89)
- SceneSplat++: A Large Dataset and Comprehensive Benchmark for Language Gaussian Splatting (NeurIPS 2025 · 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.
