Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Multi-Attention Network for One Shot Learning
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 class-tag-guided attention instead of directly pooling local visual features | One-shot image representation · Animal Dataset close-world | mAP: 67.7% -> 74.0% | Helped |
| Use class-tag-guided attention instead of directly pooling local visual features | One-shot image representation · Animal Dataset open-world | mAP: 42.1% -> 49.0% | Helped |
| Use two attention maps instead of one (ablation) | Attention network · Animal Dataset close-world | mAP: 74.0% -> 82.4% | Helped |
| Use two attention maps instead of one (ablation) | Attention network · Animal Dataset open-world | mAP: 49.0% -> 55.7% | Helped |
| Use five attention maps instead of one (ablation) | Attention network · Animal Dataset close-world | mAP: 74.0% -> 77.7% | Helped |
| Increase the attention-map count from two to five (ablation) | Attention network · Animal Dataset close-world | mAP: 82.4% -> 77.7% | Hurt |
| Increase the attention-map count from two to five (ablation) | Attention network · Animal Dataset open-world | mAP: 55.7% -> 56.8% | Helped |
| Use two attention maps instead of one (ablation) | Attention network · Artifact Dataset close-world | mAP: 34.5% -> 50.5% | Helped |
| Use two attention maps instead of one (ablation) | Attention network · Artifact Dataset open-world | mAP: 15.2% -> 28.2% | Helped |
| Apply Joint Bayesian to attention features | Attention network · Artifact Dataset close-world | mAP: 34.5% -> 36.8% | Helped |
| Apply Joint Bayesian to supervised-encoding features | Supervised encoding · Artifact Dataset open-world | mAP: 12.2% -> 11.4% | Hurt |
| Increase supervised-encoding dimensionality from 256D to 512D (ablation) | Supervised encoding · Artifact Dataset close-world | mAP: 27.8% -> 28.2% | Helped |
Closest papers by meaning
- A Dual Attention Network with Semantic Embedding for Few-Shot Learning (AAAI 2019 · similarity 0.91)
- Matching Networks for One Shot Learning (NeurIPS 2016 · similarity 0.89)
- CLEAR: Cumulative LEARning for One-Shot One-Class Image Recognition (CVPR 2018 · similarity 0.87)
- Class Semantic Attribute Perception Guided Zero-Shot Learning (AAAI 2025 · similarity 0.87)
- Pushing the Limit of Fine-Tuning for Few-Shot Learning: Where Feature Reusing Meets Cross-Scale Attention (AAAI 2024 · similarity 0.86)
- Simple Semantic-Aided Few-Shot Learning (CVPR 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.