Research map / Label-efficient and robust learning / Recognition and long-tail
Multi-Label Classification with Label Graph Superimposing
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 |
|---|---|---|---|
| combine statistical co-occurrence and knowledge graphs in the KS graph (ablation) | KSSNet · MS-COCO validation | mAP: 83.1 -> 83.7 | Helped |
| combine statistical co-occurrence and knowledge graphs in the KS graph (ablation) | KSSNet · Charades validation | mAP: 40.7 -> 44.9 | Helped |
| add statistical graph information to the knowledge graph (ablation) | KSSNet · MS-COCO validation | mAP: 81.0 -> 83.7 | Helped |
| add statistical graph information to the knowledge graph (ablation) | KSSNet · Charades validation | mAP: 41.1 -> 44.9 | Helped |
| use a knowledge graph instead of a statistical graph (ablation) | KSSNet · MS-COCO validation | mAP: 83.1 -> 81.0 | Hurt |
| use a knowledge graph instead of a statistical graph (ablation) | KSSNet · Charades validation | mAP: 40.7 -> 41.1 | Helped |
| use 4 GCN layers instead of 2 (ablation) | KSSNet · MS-COCO validation | mAP: 82.9 -> 83.7 | Helped |
| use 4 GCN layers instead of 2 (ablation) | KSSNet · Charades validation | mAP: 41.9 -> 44.9 | Helped |
| use 4 GCN layers instead of 3 (ablation) | KSSNet · MS-COCO validation | mAP: 83.5 -> 83.7 | Helped |
| use 4 GCN layers instead of 3 (ablation) | KSSNet · Charades validation | mAP: 43.8 -> 44.9 | Helped |
| add GCN label embeddings and lateral connections to the plain CNN | KSSNet · MS-COCO validation | mAP: +6.4% | Helped |
| add GCN label embeddings and lateral connections to the plain CNN | KSSNet · Charades validation | mAP: +12.0% | Helped |
Closest papers by meaning
- Multi-Label Image Recognition With Graph Convolutional Networks (CVPR 2019 · similarity 0.92)
- Enhancing the Influence of Labels on Unlabeled Nodes in Graph Convolutional Networks (ICML 2025 · similarity 0.87)
- GLDL: Graph Label Distribution Learning (AAAI 2024 · similarity 0.87)
- Multi-Label Node Classification with Label Influence Propagation (ICLR 2025 · similarity 0.86)
- Correlative and Discriminative Label Grouping for Multi-Label Visual Prompt Tuning (CVPR 2025 · similarity 0.85)
- Addressing Multi-Label Learning with Partial Labels: From Sample Selection to Label Selection (AAAI 2025 · similarity 0.85)
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.