Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
A Dual Attention Network with Semantic Embedding for Few-Shot Learning
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 semantic regularization (semantic branch) (ablation) | STANet(3-Layer) · MiniImageNet 5-way 1-shot | accuracy: 55.52 -> 58.35 | Helped |
| add semantic regularization (semantic branch) (ablation) | STANet(3-Layer) · MiniImageNet 5-way 5-shot | accuracy: 66.75 -> 71.07 | Helped |
| learned spatial attention instead of uniform attention (ablation) | STANet(3-Layer) · MiniImageNet 5-way 1-shot | accuracy: 53.41 -> 58.35 | Helped |
| learned spatial attention instead of Gaussian attention (ablation) | STANet(3-Layer) · MiniImageNet 5-way 1-shot | accuracy: 54.29 -> 58.35 | Helped |
| learned spatial attention instead of uniform attention (ablation) | STANet(3-Layer) · MiniImageNet 5-way 5-shot | accuracy: 64.32 -> 71.07 | Helped |
| learned spatial attention instead of Gaussian attention (ablation) | STANet(3-Layer) · MiniImageNet 5-way 5-shot | accuracy: 65.41 -> 71.07 | Helped |
| stack 3 STA layers instead of 1 (ablation) | STANet ResNet-12 · MiniImageNet 5-way 1-shot | accuracy: 57.25 -> 58.35 | Helped |
| stack 3 STA layers instead of 1 (ablation) | STANet ResNet-12 · MiniImageNet 5-way 5-shot | accuracy: 69.45 -> 71.07 | Helped |
| stack 3 STA layers instead of 1 (ablation) | STANet-S Conv64 · MiniImageNet 5-way 1-shot | accuracy: 50.38 -> 53.11 | Helped |
| stack 3 STA layers instead of 1 (ablation) | STANet Conv64 · Omniglot 5-way 1-shot | accuracy: 98.10 -> 98.69 | Helped |
| STANet vs MAML on Meta-CIFAR100 Easy split | STANet(3-layer) · Meta-CIFAR100 Easy 1-shot | accuracy: 55.17 -> 66.11 | Helped |
| STANet vs MAML on Meta-CIFAR100 Hard split | STANet(3-layer) · Meta-CIFAR100 Hard 1-shot | accuracy: 38.46 -> 42.98 | Helped |
Closest papers by meaning
- Pushing the Limit of Fine-Tuning for Few-Shot Learning: Where Feature Reusing Meets Cross-Scale Attention (AAAI 2024 · similarity 0.91)
- Multi-Attention Network for One Shot Learning (CVPR 2017 · similarity 0.91)
- Meta-Learning for Semi-Supervised Few-Shot Classification (ICLR 2018 · similarity 0.90)
- Dual Attention Networks for Few-Shot Fine-Grained Recognition (AAAI 2022 · similarity 0.89)
- Adaptive Cross-Modal Few-shot Learning (NeurIPS 2019 · similarity 0.89)
- One Meta-tuned Transformer is What You Need for Few-shot Learning (ICML 2024 · similarity 0.89)
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.