Research map / Graph learning / Knowledge and temporal graphs
Spatial Commonsense Graph for Object Localisation in Partial Scenes
Francesco Giuliari, Geri Skenderi, Marco Cristani, Yiming Wang, Alessio Del Bue
Check 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 commonsense AtLocation and UsedFor edges to the proximity graph (ablation) | SCG-OL · Partial scenes test set | LSR: 0.226 -> 0.238 | Improved |
| Add AtLocation edges to the proximity graph (ablation) | SCG-OL · Partial scenes test set | LSR: 0.226 -> 0.233 | Improved |
| Add UsedFor edges to the proximity graph (ablation) | SCG-OL · Partial scenes test set | LSR: 0.226 -> 0.227 | No clear effect |
| Use simultaneous rather than sequential message passing over heterogeneous edges (ablation) | SCG-OL · Partial scenes test set | LSR: 0.212 -> 0.238 | Improved |
| Use two message-passing layers rather than one (ablation) | SCG-OL · Partial scenes test set | LSR: 0.190 -> 0.238 | Improved |
| Increase message passing from two to three layers (ablation) | SCG-OL · Partial scenes test set | LSR: 0.238 -> 0.238 | No clear effect |
| Concatenate original node embeddings with contextual features instead of using contextual features alone (ablation) | SCG-OL · Partial scenes test set | LSR: 0.224 -> 0.238 | Improved |
| Use ConceptNet-pretrained node embeddings instead of learned embeddings | SCG-OL · Partial scenes test set | LSR: 0.234 -> 0.238 | Improved |
| Add commonsense knowledge to the spatial graph | GNN w/o Commonsense · Partial scenes test set | LSR: 0.212 -> 0.238 | Improved |
Closest papers by meaning
- Leveraging Commonsense for Object Localisation in Partial Scenes (TPAMI 2023 · similarity 0.96)
- SG2Loc: Sequential Visual Localization on 3D Scene Graphs (ICML 2026 · similarity 0.86)
- Multiview Scene Graph (NeurIPS 2024 · similarity 0.85)
- Learning 3D Semantic Scene Graphs From 3D Indoor Reconstructions (CVPR 2020 · similarity 0.85)
- A Comprehensive Survey of Scene Graphs: Generation and Application (TPAMI 2021 · similarity 0.85)
- GPS-Net: Graph Property Sensing Network for Scene Graph Generation (CVPR 2020 · similarity 0.84)
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.
