Research map / Graph learning / Knowledge and temporal graphs
Learning Gaussian Mixture-distributed Prototypes for 3D Scene Graph Generation from RGB-D Sequences
Rongxing Ding, Hongyu Qu, Xinguang Xiang, Pengpeng Li, Xiangbo Shu
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 GMPSSG over best prior method | 3DSSG Dense · Dense | Triplet R@1: 40.5 -> 43.0 | Improved |
| add GMPSSG over best prior method | 3DSSG GT · GT point cloud (20/8) | Triplet R@1: 67.6 -> 73.0 | Improved |
| add GMPSSG versus SCRSSG on predicate recall | 3DSSG Dense · Dense | Predicate R@1: 90.4 -> 88.3 | Hurt |
| add GMPSSG versus SCRSSG on mean object recall | 3DSSG Dense · Dense | Object mR@1: 60.5 -> 59.3 | Hurt |
| add GMP, PRL and TPI over discriminative baseline (ablation) | GMPSSG · Dense | Triplet R@1: 35.0 -> 43.0 | Improved |
| add Topology-aware Prototype Interaction (ablation) | GMP + PRL · Dense | Triplet R@1: 42.2 -> 43.0 | Improved |
| add likelihood-driven clustering loss (ablation) | GMP · Dense | Triplet R@1: 35.0 -> 41.2 | Improved |
| add cohesion-oriented loss to LLCL (ablation) | GMP · Dense | Triplet R@1: 41.2 -> 41.6 | Improved |
| add mutual information decorrelation loss (ablation) | GMP · Dense | Triplet R@1: 41.6 -> 42.2 | Improved |
| increase Gaussian components from 1 to 4 (ablation) | GMPSSG · Dense | Triplet R@1: 39.6 -> 43.0 | Improved |
| increase Gaussian components from 4 to 5 (ablation) | GMPSSG · Dense | Triplet R@1: 43.0 -> 39.0 | Hurt |
| replace standard EM with the proposed update strategy (ablation) | GMPSSG · Dense | Triplet R@1: 41.1 -> 43.0 | Improved |
Closest papers by meaning
- Bootstraping Clustering of Gaussians for View-consistent 3D Scene Understanding (AAAI 2025 · similarity 0.87)
- FHGS: Feature-Homogenized Gaussian Splatting (NeurIPS 2025 · similarity 0.87)
- Gaussian Graph Network: Learning Efficient and Generalizable Gaussian Representations from Multi-view Images (NeurIPS 2024 · similarity 0.87)
- Graph-Guided Scene Reconstruction from Images with 3D Gaussian Splatting (ICLR 2025 · similarity 0.86)
- Textured Gaussians for Enhanced 3D Scene Appearance Modeling (CVPR 2025 · similarity 0.86)
- Open3DSG: Open-Vocabulary 3D Scene Graphs from Point Clouds with Queryable Objects and Open-Set Relationships (CVPR 2024 · 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.
