5-way Few-shot Image Classification on FC100 standard (test)
80.135-way 1-shot AccuracyEfficientFSL
Evaluation results
| Method | Links | ||
|---|---|---|---|
| EfficientFSLBackbone=ViT-B, Params=2.48M, Pre-training=ImageNet21K2026.01 | 80.13 | 88.81 | |
| EfficientFSLBackbone=ViT-B, Params=2.48M, Pre-training=ImageNet1K2026.01 | 72.6 | 80.74 | |
| EfficientFSLBackbone=ViT-S, Params=1.25M, Pre-training=ImageNet1K2026.01 | 69.94 | 81.68 | |
| FewVSBackbone=ViT-S, Params=21.7M2026.01 | 61.01 | 70.37 | |
| MetaF.Backbone=ViT-S, Params=24.5M2026.01 | 58.04 | 70.8 | |
| SemFewBackbone=Swin-T, Params=88.0M2026.01 | 54.27 | 65.02 | |
| Full Fine-TuningBackbone=ViT-S, Params=21.7M2026.01 | 52.93 | 66.7 | |
| Full Fine-TuningBackbone=ViT-B, Params=85.8M2026.01 | 52.86 | 65.03 | |
| KTPPBackbone=Visformer-T, Params=11.1M2026.01 | 51.59 | 65.18 | |
| SPBackbone=Visformer-T, Params=10.3M2026.01 | 48.53 | 61.55 | |
| SCAM-NetBackbone=ViT-S, Params=21.7M2026.01 | 47.48 | 65.85 | |
| FewTUREBackbone=ViT-S, Params=21.7M2026.01 | 46.2 | 63.14 |