Generalized Few Shot Learning On Awa2
评估指标
Per-Class Accuracy (1-shot)
Per-Class Accuracy (10-shots)
Per-Class Accuracy (2-shots)
Per-Class Accuracy (5-shots)
评测结果
各个模型在此基准测试上的表现结果
| Paper Title | Repository | |||||
|---|---|---|---|---|---|---|
| MVCN | 69.9 | 82.2 | 76.4 | 81.2 | Better Generalized Few-Shot Learning Even Without Base Data | |
| CADA-VAE | 69.6 | 80.2 | 73.7 | 78.1 | Generalized Zero- and Few-Shot Learning via Aligned Variational Autoencoders | |
| DA-VAE | 68.0 | 76.8 | 73.0 | 75.6 | Generalized Zero- and Few-Shot Learning via Aligned Variational Autoencoders | - |
| DRAGON | 67.1 | 81.9 | 69.1 | 76.7 | From Generalized zero-shot learning to long-tail with class descriptors | |
| CA-VAE | 64.0 | 79.0 | 71.3 | 76.6 | Generalized Zero- and Few-Shot Learning via Aligned Variational Autoencoders | - |
| REVISE | 56.1 | 67.8 | 60.3 | 64.1 | Learning Robust Visual-Semantic Embeddings | - |
0 of 6 row(s) selected.