Research map / Label-efficient and robust learning / Recognition and long-tail
Modeling Label Correlations with Latent Context for Multi-Label Recognition
Zhaomin Chen, Quan Cui, Ruoxi Deng, Jie Hu, Guodao Zhang
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 |
|---|---|---|---|
| add multi-level context feature fusion (MLFF) to the backbone (ablation) | ResNet-101 · MS-COCO | mAP: 79.1 -> 80.9 | Improved |
| add latent context information embedding (context-aware categorical features) (ablation) | ResNet-101 + MLFF · MS-COCO | mAP: 80.9 -> 83.2 | Improved |
| add Transformer encoder for full-label correlations (ablation) | ResNet-101 + MLFF + LCIE · MS-COCO | mAP: 83.2 -> 84.0 | Improved |
| model correlations from spatial features directly instead of context-aware features (ablation) | ResNet-101 + MLFF + LCIE · MS-COCO | mAP: 82.7 vs 84.7 | Hurt |
| add partial-label correlation modeling with shared encoder (ablation) | ResNet-101 full model · MS-COCO | mAP: 84.7 -> 86.8 | Improved |
| share Transformer encoder parameters between Full and Partial (ablation) | ResNet-101 full model · MS-COCO | mAP: 85.2 -> 86.8 | Improved |
| apply the method with a stronger Swin-L (ImageNet-21K) backbone | Swin-L (21K) · MS-COCO | mAP: 89.0 -> 90.9 | Improved |
| use proposed method instead of Q2L (same input size) | multi-label recognition · MS-COCO 448x448 | mAP: 84.9 -> 86.8 | Improved |
| use proposed method with 90% of labels discarded | ResNet-101 · MS-COCO partial labels (90% discarded) | mAP: 56.6 -> 72.0 | Improved |
| use proposed method on noisy NUS-WIDE labels | ResNet-101 · NUS-WIDE | mAP: 65.1 -> 67.1 | Improved |
Closest papers by meaning
- In Pursuit of Causal Label Correlations for Multi-label Image Recognition (NeurIPS 2024 · similarity 0.90)
- Label Embedding with Partial Heterogeneous Contexts (AAAI 2019 · similarity 0.89)
- Improving Multi-label Learning with Missing Labels by Structured Semantic Correlations (ECCV 2016 · similarity 0.89)
- Collaboration Based Multi-Label Learning (AAAI 2019 · similarity 0.88)
- Correlative and Discriminative Label Grouping for Multi-Label Visual Prompt Tuning (CVPR 2025 · similarity 0.86)
- Attention-Induced Embedding Imputation for Incomplete Multi-View Partial Multi-Label Classification (AAAI 2024 · similarity 0.86)
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.
