Research map / Label-efficient and robust learning / Recognition and long-tail
Semantic-Aware Representation Blending for Multi-Label Image Recognition with Partial Labels
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 |
|---|---|---|---|
| Add SARB category-specific representation blending (ablation) | SSGRL · MS-COCO | Avg. mAP: 3.8% | Helped |
| Add SARB category-specific representation blending (ablation) | SSGRL · VG-200 | Avg. mAP: 5.9% | Helped |
| Add SARB category-specific representation blending (ablation) | SSGRL · Pascal VOC 2007 | Avg. mAP: 1.2% | Helped |
| Use SD instead of SAM for CSRL (ablation) | SARB · MS-COCO | Avg. mAP: 0.3% | Helped |
| Use SD instead of SAM for CSRL (ablation) | SARB · VG-200 | Avg. mAP: 0.2% | Helped |
| Use SD instead of SAM for CSRL (ablation) | SARB · Pascal VOC 2007 | Avg. mAP: 0.1% | Helped |
| Replace category-specific representation blending with pixel-level mixup (ablation) | SARB · MS-COCO | Avg. mAP: 3.6% | Hurt |
| Replace category-specific representation blending with feature-map mixup (ablation) | SARB · VG-200 | Avg. mAP: 6.0% | Hurt |
| Add instance-level representation blending (ablation) | SSGRL · MS-COCO | Avg. mAP: 3.2% | Helped |
| Learn α instead of fixing it to 0.5 (ablation) | ILRB · VG-200 | Avg. mAP: 44.5% to 44.9% | Helped |
| Add prototype-level representation blending (ablation) | SSGRL · VG-200 | Avg. mAP: 5.2% | Helped |
| Learn β instead of fixing it to 0.5 (ablation) | PLRB · MS-COCO | Avg. mAP: 0.4% | Helped |
Closest papers by meaning
- Structured Semantic Transfer for Multi-Label Recognition with Partial Labels (AAAI 2022 · similarity 0.91)
- [CLS] is Not Enough: Multi-Label Recognition via Patch-Level Inference and Adaptive Aggregation (ICML 2026 · similarity 0.90)
- Incomplete Multi-View Multi-label Learning via Disentangled Representation and Label Semantic Embedding (CVPR 2025 · similarity 0.88)
- Partial Multi-View Multi-Label Classification via Semantic Invariance Learning and Prototype Modeling (ICML 2024 · similarity 0.87)
- Correlative and Discriminative Label Grouping for Multi-Label Visual Prompt Tuning (CVPR 2025 · similarity 0.87)
- Navigating Real-World Partial Label Learning: Unveiling Fine-Grained Images with Attributes (AAAI 2024 · 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.