Table 4 72-way classification accuracy on SUN of proposed model
From: Multimodal few-shot classification without attribute embedding
Backbone | Metric | 1-shot | 2-shot | 5-shot | 10-shot |
|---|---|---|---|---|---|
ResNet-18 | Top-1 | 10.70 | 23.52 | 49.91 | 62.95 |
Top-3 | 17.61 | 39.69 | 70.58 | 83.29 | |
Top-5 | 20.21 | 47.63 | 80.30 | 90.03 | |
ResNet-101 | Top-1 | 36.40 | 46.14 | 59.26 | 66.67 |
Top-3 | 55.41 | 67.28 | 80.56 | 87.08 | |
Top-5 | 65.20 | 76.23 | 87.31 | 94.31 |