Table 4 Proposed accuracy methods using different folds
From: Facial expression recognition using three-stage support vector machines
Datasets | Methods | |||
|---|---|---|---|---|
Leave-one-out | 10 folds | 5 folds | 2 folds | |
JAFFE | 96.71% | 98.10% | 97.62% | 90.10% |
CK+ | 93.29% | 94.42% | 93.49% | 90.00% |
RaFD | 99.72% | 95.14% | 95.10% | 94.88% |