Research map / Graph learning / Knowledge and temporal graphs
Learning Physical Graph Representations from Visual Scenes
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 PSGNet bottom-up graph grouping instead of latent-factor CNN baselines (MONet, IODINE) | PSGNetS · Primitives | epochs to peak validation recall: 200 -> 2.0 (MONet); 350 -> 2.0 (IODINE) | Helped |
| use PSGNet instead of CNN baselines | PSGNetM · Playroom | epochs to peak validation recall: 38 -> 1.0 (MONet); 24 -> 1.0 (OP3) | Helped |
| feed depth and normal maps as inputs, or reconstruct RGB only (ablation) | MONet and IODINE · Primitives | segmentation performance | No clear effect |
| add motion grouping (P3, P4) to PSGNetS (PSGNetM) | PSGNet · Playroom | grouping errors | Mixed |
Closest papers by meaning
- Neural Scene Graphs for Dynamic Scenes (CVPR 2021 · similarity 0.89)
- Object-Centric Representation Learning for Enhanced 3D Semantic Scene Graph Prediction (NeurIPS 2025 · similarity 0.89)
- Adaptive Visual Scene Understanding: Incremental Scene Graph Generation (NeurIPS 2024 · similarity 0.88)
- SCENIR: Visual Semantic Clarity through Unsupervised Scene Graph Retrieval (ICML 2025 · similarity 0.87)
- Multi-Part Object Representations via Graph Structures and Co-Part Discovery (NeurIPS 2026 · similarity 0.87)
- 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.