Research map / Graph learning / Knowledge and temporal graphs
Graph Edit Distance Reward: Learning to Edit Scene Graph
Lichang Chen, Guosheng Lin, Shijie Wang, Qingyao Wu
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 |
|---|---|---|---|
| use GED reward instead of 0,1 reward for finetuning (ablation) | Program Generator · CRIR (overall) | Recall@1 (%): 95.4 -> 98.4 | Improved |
| use GED reward instead of 0,1 reward for finetuning (ablation) | Program Generator · CRIR (Zero Hop) | Recall@1 (%): 95.9 -> 99.0 | Improved |
| use GED reward instead of 0,1 reward for finetuning (ablation) | Program Generator · CRIR (One Hop) | Recall@1 (%): 95.4 -> 98.4 | Improved |
| use GED reward instead of 0,1 reward for finetuning (ablation) | Program Generator · CRIR (Two Hop) | Recall@1 (%): 95.8 -> 98.2 | Improved |
| use GED reward instead of 0,1 reward for finetuning (ablation) | Program Generator · CRIR (Three Hop) | Recall@1 (%): 95.3 -> 97.9 | Improved |
| use GED reward instead of 0,1 reward for finetuning (ablation) | Program Generator · CRIR (Single And) | Recall@1 (%): 94.5 -> 98.4 | Improved |
| replace feature-combination TIRG with scene graph editing + GED reward | TIRG · CSS (overall) | Recall@1 (%): 73.7 -> 99.8 | Improved |
| replace feature-combination TIRG with scene graph editing + GED reward | TIRG · CRIR (overall) | Recall@1 (%): 26.1 -> 98.4 | Improved |
| replace feature-combination TIRG with scene graph editing + GED reward | TIRG · CRIR (Three Hop) | Recall@1 (%): 4.8 -> 97.9 | Improved |
| replace feature-combination TIRG with scene graph editing + GED reward | TIRG · CSS (Remove queries) | Recall@1 (%): 64.3 -> 99.8 | Improved |
Closest papers by meaning
- SCENIR: Visual Semantic Clarity through Unsupervised Scene Graph Retrieval (ICML 2025 · similarity 0.85)
- Position: Graph Matching Systems Deserve Better Benchmarks (ICML 2025 · similarity 0.84)
- Graph Edit Distance with General Costs Using Neural Set Divergence (NeurIPS 2024 · similarity 0.84)
- Scene Graph Disentanglement and Composition for Generalizable Complex Image Generation (NeurIPS 2024 · similarity 0.84)
- Gradient Rewiring for Editable Graph Neural Network Training (NeurIPS 2024 · similarity 0.83)
- Mosaic of Modalities: A Comprehensive Benchmark for Multimodal Graph Learning (CVPR 2025 · similarity 0.83)
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.
