Research map / Label-efficient and robust learning / Recognition and long-tail
Label Embedding with Partial Heterogeneous Contexts
Yaxin Shi, Donna Xu, Yuangang Pan, Ivor W. Tsang, Shirui Pan
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 PHCLE instead of the strongest single-context or fusion baseline | zero-shot classification embeddings · AWA, ESZSL | average per-class top-1 accuracy: 63.66 -> 69.11 | Improved |
| use PHCLE instead of the strongest baseline | zero-shot classification embeddings · AWA, ConSE | average per-class top-1 accuracy: 58.08 -> 58.43 | Improved |
| use PHCLE instead of the strongest baseline | zero-shot classification embeddings · AWA, SJE | average per-class top-1 accuracy: 74.27 -> 77.47 | Improved |
| use PHCLE instead of the strongest baseline | zero-shot classification embeddings · aPY, ESZSL | average per-class top-1 accuracy: 52.60 -> 54.61 | Improved |
| use PHCLE instead of the strongest baseline | zero-shot classification embeddings · aPY, ConSE | average per-class top-1 accuracy: 52.87 -> 52.87 | No clear effect |
| use PHCLE instead of the strongest baseline | zero-shot classification embeddings · aPY, SJE | average per-class top-1 accuracy: 52.18 -> 50.91 | Hurt |
| use PHCLE instead of TADW with full context correspondence | full-context label embeddings · AWA, ESZSL | average per-class top-1 accuracy: 50.18 -> 69.08 | Improved |
| use PHCLE instead of TADW with full context correspondence | full-context label embeddings · AWA, ConSE | average per-class top-1 accuracy: 56.62 -> 56.77 | Improved |
| use PHCLE instead of TADW with full context correspondence | full-context label embeddings · AWA, SJE | average per-class top-1 accuracy: 40.84 -> 53.33 | Improved |
| use PHCLE instead of TADW with full context correspondence | full-context label embeddings · aPY, ESZSL | average per-class top-1 accuracy: 28.13 -> 50.56 | Improved |
| use PHCLE instead of TADW with full context correspondence | full-context label embeddings · aPY, ConSE | average per-class top-1 accuracy: 25.21 -> 53.04 | Improved |
| use PHCLE instead of TADW with full context correspondence | full-context label embeddings · aPY, SJE | average per-class top-1 accuracy: 23.48 -> 42.52 | Improved |
Closest papers by meaning
- Modeling Label Correlations with Latent Context for Multi-Label Recognition (ECCV 2024 · similarity 0.89)
- Heterogeneous Semantic Transfer for Multi-label Recognition with Partial Labels (IJCV 2024 · similarity 0.86)
- Multi-View Multi-Instance Multi-Label Learning Based on Collaborative Matrix Factorization (AAAI 2019 · similarity 0.86)
- Navigating Real-World Partial Label Learning: Unveiling Fine-Grained Images with Attributes (AAAI 2024 · similarity 0.85)
- Incomplete Multi-View Multi-label Learning via Disentangled Representation and Label Semantic Embedding (CVPR 2025 · similarity 0.85)
- Learning with Partial-Label and Unlabeled Data: A Uniform Treatment for Supervision Redundancy and Insufficiency (ICML 2024 · 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.
