Table 2 Experimental results for S2S and S2R in the second 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 2: S2S | Session 2: S2R | |||||||||
Acc. | Pre. | Sens. | Spec. | F1 | Acc. | Pre. | Sens. | Spec. | F1 | |
GoogLeNet | 99.19 ± 0.53 | 99.21 ± 0.53 | 99.17 ± 0.54 | 99.96 ± 0.02 | 99.18 ± 0.55 | 99.42 ± 0.23 | 99.42 ± 0.22 | 99.39 ± 0.24 | 99.96 ± 0.01 | 99.40 ± 0.24 |
Inception-V3 | 99.84 ± 0.23 | 99.85 ± 0.21 | 99.84 ± 0.23 | 99.98 ± 0.01 | 99.84 ± 0.23 | 99.95 ± 0.08 | 99.95 ± 0.28 | 99.94 ± 0.08 | 99.98 ± 0.003 | 99.94 ± 0.08 |
ResNet-50 | 99.88 ± 0.01 | 99.90 ± 0.002 | 99.90 ± 0.01 | 99.99 ± 0.02 | 99.89 ± 0.001 | 99.89 ± 0.002 | 99.90 ± 0.01 | 99.44 ± 0.63 | 99.99 ± 0.02 | 99.89 ± 0.002 |
ResNet-101 | 99.68 ± 0.15 | 99.69 ± 0.14 | 99.66 ± 0.14 | 99.97 ± 0.01 | 99.67 ± 0.14 | 99.79 ± 0.30 | 99.78 ± 0.31 | 99.78 ± 0.31 | 99.98 ± 0.01 | 99.78 ± 0.31 |
EfficientNet-B0 | 99.89 ± 0.16 | 99.87 ± 0.019 | 99.86 ± 0.20 | 99.98 ± 0.01 | 99.86 ± 0.19 | 99.84 ± 0.23 | 99.84 ± 0.23 | 99.83 ± 0.24 | 99.98 ± 0.01 | 99.83 ± 0.24 |
DenseNet-201 | 99.98 ± 0.01 | 99.98 ± 0.02 | 99.98 ± 0.02 | 99.99 ± 0.01 | 99.99 ± 0.01 | 99.89 ± 0.01 | 99.90 ± 0.01 | 99.88 ± 0.01 | 99.99 ± 0.02 | 99.89 ± 0.01 |