Research map / Point clouds / Point cloud classification
3D Spatial Multimodal Knowledge Accumulation for Scene Graph Prediction in Point Cloud
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| use 3D spatial multimodal knowledge accumulation | 3D scene graph prediction · 3DSSG PredCls | R@50: 64.47 -> 68.32 | Helped |
| use 3D spatial multimodal knowledge accumulation | 3D scene graph prediction · 3DSSG SGCls | mR@50: 28.20 -> 30.29 | Helped |
| add hierarchical tokens to symbolic knowledge (ablation) | 3D spatial scene graph model · 3DSSG SGCls | R@50: 30.47 -> 31.50 | Helped |
| add support edges to symbolic knowledge (ablation) | 3D spatial scene graph model · 3DSSG SGCls | R@50: 30.55 -> 31.50 | Helped |
| add hierarchical tokens and support edges (ablation) | 3D spatial scene graph model · 3DSSG SGCls | R@50: 28.41 -> 31.50 | Helped |
| use a hierarchical visual graph instead of a fully connected graph (ablation) | 3D spatial scene graph model · 3DSSG SGCls | R@50: 28.17 -> 31.50 | Helped |
| add the region-aware graph network (ablation) | 3D spatial scene graph model · 3DSSG SGCls | R@50: 26.43 -> 31.50 | Helped |
| use a region-aware graph network rather than a standard GCN (ablation) | 3D spatial scene graph model · 3DSSG SGCls | R@50: 31.03 -> 31.50 | Helped |
| add multimodal knowledge embeddings (ablation) | 3D spatial scene graph model · 3DSSG SGCls | mR@50: 22.93 -> 30.29 | Helped |
| use visual contextual features as graph reasoning inputs (ablation) | 3D spatial scene graph model · 3DSSG SGCls | R@50: 28.14 -> 31.50 | Helped |
| use a ground-truth support graph instead of a randomly connected graph (ablation) | visual graph construction · 3DSSG PredCls | R@50: 62.74 -> 68.41 | Helped |
| use a hierarchical visual graph instead of a randomly connected graph (ablation) | visual graph construction · 3DSSG SGCls | R@50: 28.17 -> 31.50 | Helped |
Closest papers by meaning
- Open3DSG: Open-Vocabulary 3D Scene Graphs from Point Clouds with Queryable Objects and Open-Set Relationships (CVPR 2024 · similarity 0.90)
- Object-Centric Representation Learning for Enhanced 3D Semantic Scene Graph Prediction (NeurIPS 2025 · similarity 0.90)
- Open-Vocabulary Functional 3D Scene Graphs for Real-World Indoor Spaces (CVPR 2025 · similarity 0.88)
- SceneScaffold: Active Scene-State Construction for Unified 3D Scene Understanding (NeurIPS 2026 · similarity 0.88)
- Scene Graph Thinking: Reinforcing Structured Visual Reasoning for Multimodal Large Language Models (ICML 2026 · similarity 0.88)
- SGFormer: Semantic Graph Transformer for Point Cloud-Based 3D Scene Graph Generation (AAAI 2024 · 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.