Research map / 3D vision / Gaussian splatting
Trace3D: Consistent Segmentation Lifting via Gaussian Instance Tracing
Hongyu Shen, Junfeng Ni, Yixin Chen, Weishuo Li, Mingtao Pei, Siyuan Huang
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 GIT-guided ambiguous Gaussian control (ablation) | Trace3D on 2DGS, no consistent mask · Replica 3D object extraction | object mIoU: 62.5 -> 70.1 | Improved |
| add GIT-guided ambiguous Gaussian control (ablation) | Trace3D on 3DGS · Replica 3D object extraction | object mIoU: 60.9 -> 63.7 | Improved |
| add GIT-guided ambiguous Gaussian control on top of consistent 2D mask (ablation) | Trace3D on 2DGS · Replica 3D object extraction | object mIoU: 63.6 -> 72.1 | Improved |
| add consistent 2D mask (ablation) | Trace3D on 2DGS, no density control · Replica 3D object extraction | object mIoU: 62.5 -> 63.6 | Improved |
| add consistent 2D mask on top of density control (ablation) | Trace3D on 2DGS · Replica 3D object extraction | object mIoU: 70.1 -> 72.1 | Improved |
| add consistent 2D mask (ablation) | Trace3D on 2DGS, no density control · Replica novel view 2D segmentation | NVS mIoU: 87.0 -> 89.2 | Improved |
| add ambiguous Gaussian control without consistent mask (ablation) | Trace3D on 2DGS · Replica novel view 2D segmentation | NVS mIoU: 87.0 -> 87.3 | No clear effect |
| use contrastive lifting instead of self-prompting (ablation) | Trace3D on 2DGS · Replica novel view 2D segmentation | NVS mIoU: 72.7 -> 89.1 | Improved |
| use contrastive lifting instead of self-prompting (ablation) | Trace3D on 2DGS · Replica 3D object extraction | object mIoU: 57.7 -> 72.1 | Improved |
| use 2DGS instead of 3DGS representation with full method (ablation) | Trace3D · Replica 3D object extraction | object mIoU: 63.7 -> 72.1 | Improved |
| add control plus consistent mask to 3DGS PSNR (ablation) | Trace3D on 3DGS · Replica 3D object extraction | PSNR: 20.8 -> 21.2 | Improved |
| add ambiguous Gaussian control without mask (ablation) | Trace3D on 2DGS · Replica 3D object extraction | PSNR: 21.0 -> 22.4 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 4 on most shared results.
- FlashSplat — FlashSplat: 2D to 3D Gaussian Splatting Segmentation Solved Optimally (ECCV 2024) · better on 3 of 4 shared results
- OmniSeg3D — OmniSeg3D: Omniversal 3D Segmentation via Hierarchical Contrastive Learning (CVPR 2024) · better on 3 of 3 shared results
- GaussianCut — GaussianCut: Interactive segmentation via graph cut for 3D Gaussian Splatting (NeurIPS 2024) · better on 1 of 2 shared results
- Gaussian Grouping — Gaussian Grouping: Segment and Edit Anything in 3D Scenes (ECCV 2024) · better on 2 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
- Rethinking End-to-End 2D to 3D Scene Segmentation in Gaussian Splatting (CVPR 2025 · similarity 0.89)
- GaussianCut: Interactive segmentation via graph cut for 3D Gaussian Splatting (NeurIPS 2024 · similarity 0.89)
- Bootstraping Clustering of Gaussians for View-consistent 3D Scene Understanding (AAAI 2025 · similarity 0.86)
- ReferSplat: Referring Segmentation in 3D Gaussian Splatting (ICML 2025 · similarity 0.86)
- InstanceGaussian: Appearance-Semantic Joint Gaussian Representation for 3D Instance-Level Perception (CVPR 2025 · similarity 0.86)
- Segment then Splat: Unified 3D Open-Vocabulary Segmentation via Gaussian Splatting (NeurIPS 2025 · similarity 0.86)
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.
