Research map / Graph learning / Knowledge and temporal graphs
Multi-Label Image Recognition With Graph Convolutional Networks
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 ML-GCN with a re-weighted correlation matrix instead of the strongest reported baseline | MS-COCO recognition · MS-COCO val | mAP: 77.3 -> 83.0 | Helped |
| use ML-GCN with a re-weighted correlation matrix instead of the strongest reported baseline | MS-COCO recognition · MS-COCO val | OF1: 78.4 -> 80.3 | Helped |
| use ML-GCN with a re-weighted correlation matrix instead of the strongest reported baseline | VOC 2007 recognition · VOC 2007 test | mAP: 92.0 -> 94.0 | Helped |
| replace the binary correlation matrix with the re-weighted correlation matrix (ablation) | ML-GCN · MS-COCO val | mAP: 80.3 -> 83.0 | Helped |
| replace the binary correlation matrix with the re-weighted correlation matrix (ablation) | ML-GCN · VOC 2007 test | mAP: 93.1 -> 94.0 | Helped |
| set the neighbor weight p to 0, removing explicit label-correlation propagation (ablation) | ML-GCN · MS-COCO val | mAP: 81.67% compared with the best performance at p=0.2 | Hurt |
| set the neighbor weight p to 0, removing explicit label-correlation propagation (ablation) | ML-GCN · VOC 2007 test | mAP: 93.15% compared with the best performance at p=0.2 | Hurt |
| increase GCN depth from 2 to 3 layers (ablation) | ML-GCN · MS-COCO val | mAP: 83.0 -> 82.1 | Hurt |
| increase GCN depth from 2 to 4 layers (ablation) | ML-GCN · MS-COCO val | mAP: 83.0 -> 81.1 | Hurt |
| increase GCN depth from 2 to 3 layers (ablation) | ML-GCN · VOC 2007 test | mAP: 94.0 -> 93.6 | Hurt |
| increase GCN depth from 2 to 4 layers (ablation) | ML-GCN · VOC 2007 test | mAP: 94.0 -> 93.0 | Hurt |
| filter low-probability label edges using a threshold (ablation) | ML-GCN correlation graph · MS-COCO and VOC 2007 | recognition accuracy | Mixed |
Closest papers by meaning
- Modular Graph Transformer Networks for Multi-Label Image Classification (AAAI 2021 · similarity 0.92)
- Multi-Label Classification with Label Graph Superimposing (AAAI 2020 · similarity 0.92)
- CNN-RNN: A Unified Framework for Multi-Label Image Classification (CVPR 2016 · similarity 0.91)
- GLDL: Graph Label Distribution Learning (AAAI 2024 · similarity 0.89)
- Enhancing the Influence of Labels on Unlabeled Nodes in Graph Convolutional Networks (ICML 2025 · similarity 0.87)
- Correlative and Discriminative Label Grouping for Multi-Label Visual Prompt Tuning (CVPR 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.