Table 1 Experimental results for S1S and S1R in the first session based on different pre-trained models
From: An evaluation of transfer learning models in EEG-based authentication
Model | Comparison of performance metrics (Averaged% ± Standard deviation) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
Session 1: S1S | Session 1: S1R | |||||||||
Acc. | Pre. | Sens. | Spec. | F1 | Acc. | Pre. | Sens. | Spec. | F1 | |
GoogLeNet | 99.79 ± 0.01 | 99.80 ± 0.002 | 99.79 ± 0.005 | 99.99 ± 0.001 | 99.79 ± 0.001 | 99.63 ± 0.08 | 99.63 ± 0.10 | 99.61 ± 0.07 | 99.97 ± 0.003 | 99.61 ± 0.08 |
Inception-V3 | 99.95 ± 0.08 | 99.95 ± 0.08 | 99.94 ± 0.08 | 99.99 ± 0.002 | 99.94 ± 0.08 | 99.95 ± 0.08 | 99.95 ± 0.07 | 99.94 ± 0.09 | 99.99 ± 0.002 | 99.94 ± 0.08 |
ResNet-50 | 99.98 ± 0.01 | 99.98 ± 0.01 | 99.98 ± 0.01 | 99.99 ± 0.002 | 99.98 ± 0.002 | 99.89 ± 0.002 | 99.90 ± 0.01 | 99.88 ± 0.002 | 99.98 ± 0.01 | 99.88 ± 0.003 |
ResNet-101 | 99.95 ± 0.08 | 99.95 ± 0.07 | 99.94 ± 0.09 | 99.99 ± 0.003 | 99.94 ± 0.08 | 99.84 ± 0.23 | 99.84 ± 0.22 | 99.82 ± 0.25 | 99.98 ± 0.002 | 99.83 ± 0.24 |
EfficientNet-B0 | 99.84 ± 0.08 | 99.84 ± 0.07 | 99.83 ± 0.09 | 99.99 ± 0.003 | 99.83 ± 0.08 | 99.89 ± 0.002 | 99.89 ± 0.01 | 99.89 ± 0.01 | 99.98 ± 0.002 | 99.89 ± 0.001 |
DenseNet-201 | 99.98 ± 0.01 | 99.98 ± 0.01 | 99.98 ± 0.01 | 99.99 ± 0.002 | 99.98 ± 0.01 | 99.95 ± 0.08 | 99.95 ± 0.07 | 99.94 ± 0.09 | 99.99 ± 0.003 | 99.94 ± 0.08 |