Research map / Graph learning / Knowledge and temporal graphs
Exploiting Edge-Oriented Reasoning for 3D Point-Based Scene Graph Analysis
Check 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 EdgeGCN with explicit edge features and twinning attention instead of the strongest reported alternatives | PointNet scene-graph reasoning · 3RScan | predicate F1@3: 72.57 -> 78.88 | Helped |
| Use EdgeGCN instead of PointNet without graph reasoning | PointNet scene-graph reasoning · 3RScan | relationship triplet R@100: 45.86 -> 48.68 | Helped |
| Use EdgeGCN instead of the best listed non-missing-object baseline | SGGpoint · SUNCG Living | accuracy: 74.7 -> 76.2 | Helped |
| Use EdgeGCN instead of the best listed non-missing-object baseline | SGGpoint · SUNCG Office | accuracy: 73.0 -> 74.1 | Helped |
| Use EdgeGCN instead of the best listed non-missing-object baseline | SGGpoint · SUNCG Bed | accuracy: 79.9 -> 79.5 | Hurt |
| Use EdgeGCN instead of the best listed non-missing-object baseline | SGGpoint · SUNCG Bath | accuracy: 56.4 -> 61.6 | Helped |
| Add EdgeGCN edge-driven reasoning | EdgeGCN on citation and molecular graphs · Cora | accuracy: 80.3 -> 81.6 | Helped |
| Add EdgeGCN edge-driven reasoning | EdgeGCN on citation and molecular graphs · CiteSeer | accuracy: 67.7 -> 69.4 | Helped |
| Add EdgeGCN edge-driven reasoning | EdgeGCN on citation and molecular graphs · Pubmed | accuracy: 78.5 -> 78.7 | Mixed |
| Add edge-driven attention to Vanilla EdgeGCN (ablation) | EdgeGCN scene-graph reasoning · 3RScan ablation (PointNet and DGCNN) | node R@1: 54.8 -> 56.9; 60.8 -> 61.7 | Helped |
| Add edge-driven attention to Vanilla EdgeGCN (ablation) | EdgeGCN scene-graph reasoning · 3RScan ablation (PointNet and DGCNN) | edge F1@1: 41.9 -> 41.1; 43.9 -> 41.1 | Hurt |
| Add node-driven attention to Vanilla EdgeGCN (ablation) | EdgeGCN scene-graph reasoning · 3RScan ablation (PointNet and DGCNN) | edge F1@1: 41.9 -> 50.0; 43.9 -> 47.4 | Helped |
Closest papers by meaning
- Learning 3D Semantic Scene Graphs From 3D Indoor Reconstructions (CVPR 2020 · similarity 0.91)
- Edge-Centric Relational Reasoning for 3D Scene Graph Prediction (AAAI 2026 · similarity 0.90)
- Scene Graph Generation by Iterative Message Passing (CVPR 2017 · similarity 0.90)
- SGFormer: Semantic Graph Transformer for Point Cloud-Based 3D Scene Graph Generation (AAAI 2024 · similarity 0.88)
- Scene Graph-Grounded Image Generation (AAAI 2025 · similarity 0.88)
- DSGG: Dense Relation Transformer for an End-to-end Scene Graph Generation (CVPR 2024 · similarity 0.87)
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.