Research map / Graph learning / Graph algorithms and clustering
Graph-Context Attention Networks for Size-Varied Deep Graph Matching
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 GCAN with ILP attention loss instead of the strongest listed baseline | Pascal VOC Keypoint matching with intersection filtering · Pascal VOC Keypoint, intersection filtering | mean accuracy: 80.4 -> 82.3 | Helped |
| Use GCAN for size-varied graph matching instead of BBGM | Pascal VOC Keypoint matching · Pascal VOC Keypoint, size-varied graphs | mean F1: 61.4 -> 64.6 | Helped |
| Train GCAN with ILP attention loss rather than binary cross-entropy (ablation) | GCAN · Pascal VOC Keypoint, intersection filtering | mean accuracy: 81.3 -> 82.3 | Helped |
| Train GCAN with ILP attention loss rather than Hungarian Attention Loss (ablation) | GCAN · Pascal VOC Keypoint, intersection filtering | mean accuracy: 81.2 -> 82.3 | Helped |
| Train GCAN with ILP attention loss rather than pixel offset regression loss (ablation) | GCAN · Pascal VOC Keypoint, intersection filtering | mean accuracy: 78.7 -> 82.3 | Helped |
| Train GCAN with ILP attention loss rather than focal loss (ablation) | GCAN · Pascal VOC Keypoint, intersection filtering | mean accuracy: 77.7 -> 82.3 | Helped |
| Train GCAN with ILP attention loss rather than cost margin loss (ablation) | GCAN · Pascal VOC Keypoint, intersection filtering | mean accuracy: 40.2 -> 82.3 | Helped |
| Add cross-attention to the baseline model (ablation) | GCAN architecture · Pascal VOC Keypoint | accuracy: 58.1 -> 70.2 | Helped |
| Add positional encoding to the cross-attention model (ablation) | GCAN architecture · Pascal VOC Keypoint | accuracy: 70.2 -> 79.4 | Helped |
| Add self-attention to the model with cross-attention and positional encoding (ablation) | GCAN architecture · Pascal VOC Keypoint | accuracy: 79.4 -> 82.3 | Helped |
Closest papers by meaning
- Learning deep graph matching with channel-independent embedding and Hungarian attention (ICLR 2020 · similarity 0.92)
- Deep Learning of Graph Matching (CVPR 2018 · similarity 0.90)
- ClusterGNN: Cluster-Based Coarse-To-Fine Graph Neural Network for Efficient Feature Matching (CVPR 2022 · similarity 0.88)
- Deep Learning of Partial Graph Matching via Differentiable Top-K (CVPR 2023 · similarity 0.88)
- ResMatch: Residual Attention Learning for Feature Matching (AAAI 2024 · similarity 0.87)
- Learning Partial Graph Matching via Optimal Partial Transport (ICLR 2025 · 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.