Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Discriminative Region-Based Multi-Label Zero-Shot Learning
Sanath Narayan, Akshita Gupta, Salman Khan, Fahad Shahbaz Khan, Ling Shao, Mubarak Shah
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 region-level classification before spatial pooling instead of pooling features first (ablation) | BiAM · NUS-WIDE ZSL/GZSL | mAP | Improved |
| add RCB to standard region features (ablation) | region-level framework · NUS-WIDE ZSL | mAP: 21.1 -> 23.7 | Improved |
| add SCB to standard region features (ablation) | region-level framework · NUS-WIDE ZSL | mAP: 21.1 -> 23.2 | Improved |
| add SCB to RCB (full BiAM) (ablation) | BiAM · NUS-WIDE ZSL | mAP: 23.7 -> 26.3 | Improved |
| add RCB to SCB (full BiAM) (ablation) | BiAM · NUS-WIDE ZSL | mAP: 23.2 -> 26.3 | Improved |
| add RCB to standard region features (ablation) | region-level framework · NUS-WIDE GZSL | mAP: 6.8 -> 7.6 | Improved |
| add SCB to standard region features (ablation) | region-level framework · NUS-WIDE GZSL | mAP: 6.8 -> 8.6 | Improved |
| use full BiAM instead of standard region features (ablation) | region-level framework · NUS-WIDE GZSL | mAP: 6.8 -> 9.3 | Improved |
| add LayerNorm in RCB (ablation) | BiAM · NUS-WIDE ZSL | mAP: 26.3 -> 25.0 | Hurt |
| replace softmax with sigmoid in RCB (ablation) | BiAM · NUS-WIDE ZSL | mAP: 26.3 -> 24.6 | Hurt |
| replace sigmoid with softmax in SCB (ablation) | BiAM · NUS-WIDE ZSL | mAP: 26.3 -> 24.3 | Hurt |
| replace Non-Local attention with BiAM (ablation) | region-level framework · NUS-WIDE ZSL | mAP: 23.1 -> 26.3 | Improved |
Closest papers by meaning
- Epsilon: Exploring Comprehensive Visual-Semantic Projection for Multi-Label Zero-Shot Learning (AAAI 2025 · similarity 0.92)
- A Shared Multi-Attention Framework for Multi-Label Zero-Shot Learning (CVPR 2020 · similarity 0.90)
- Rethinking Zero-Shot Learning: A Conditional Visual Classification Perspective (ICCV 2019 · similarity 0.90)
- Discriminative Learning of Latent Features for Zero-Shot Recognition (CVPR 2018 · similarity 0.90)
- ZeroDiff: Solidified Visual-semantic Correlation in Zero-Shot Learning (ICLR 2025 · similarity 0.89)
- TPR: Topology-Preserving Reservoirs for Generalized Zero-Shot Learning (NeurIPS 2024 · similarity 0.88)
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.
