Generalized Few-Shot Learning 벤치마크
Generalized Few-Shot Learning on AwA2
Per-Class Accuracy (1-shot)
- 2017-03-17 — REVISE: Per-Class Accuracy (1-shot) 56.1
- 2018-12-05 — CADA-VAE: Per-Class Accuracy (1-shot) 69.6
- 2022-11-29 — MVCN: Per-Class Accuracy (1-shot) 69.9
| Rank | Model | Per-Class Accuracy (1-shot) | Per-Class Accuracy (2-shots) | Per-Class Accuracy (5-shots) | Per-Class Accuracy (10-shots) | Per-Class Accuracy (20-shots) | Paper | Code | Year |
|---|---|---|---|---|---|---|---|---|---|
| 1 | MVCN | 69.9 | 76.4 | 81.2 | 82.2 | – | Better Generalized Few-Shot Learning Even Without Base Data | bigdata-inha/zero-base-gfsl | 2022 |
| 2 | CADA-VAE | 69.6 | 73.7 | 78.1 | 80.2 | 80.9 | Generalized Zero- and Few-Shot Learning via Aligned Variational Autoencoders | edgarschnfld/CADA-VAE-PyTorch · sanixa/CADA-VAE-pytorch | 2018 |
| 3 | DA-VAE | 68.0 | 73.0 | 75.6 | 76.8 | – | Generalized Zero- and Few-Shot Learning via Aligned Variational Autoencoders | edgarschnfld/CADA-VAE-PyTorch | 2019 |
| 4 | DRAGON | 67.1 | 69.1 | 76.7 | 81.9 | 83.3 | From Generalized zero-shot learning to long-tail with class descriptors | dvirsamuel/DRAGON | 2020 |
| 5 | CA-VAE | 64.0 | 71.3 | 76.6 | 79.0 | – | Generalized Zero- and Few-Shot Learning via Aligned Variational Autoencoders | edgarschnfld/CADA-VAE-PyTorch | 2019 |
| 6 | REVISE | 56.1 | 60.3 | 64.1 | 67.8 | – | Learning Robust Visual-Semantic Embeddings | 2017 |