Table 5 Small-sample robustness evaluation of ViT-HVE on the CHVR dataset

From: ViT-HVE: a vision transformer-based framework for recognition and weighted evaluation of cultural heritage values

Setting

Accuracy

1-shot

51.21

5-shot

58.82

10-shot

63.45

20-shot

69.73

Full data

88.90

  1. We report Top-1 Accuracy (%) under different K-shot settings (K = 1, 5, 10, 20).
Back to article page