Table 3.
The subject-specific session to session classification results (accuracy in percentage) on dataset 2b. The best accuracies are highlighted in bold; the lowest accuracies of SCNN-MCD with respect to SCNN are underlined.
| Methods | B01 | B02 | B03 | B04 | B05 | B06 | B07 | B08 | B09 | Mean |
|---|---|---|---|---|---|---|---|---|---|---|
| CNN-SAE [35] | 78.10 | 63.10 | 60.60 | 95.60 | 78.10 | 73.80 | 70.00 | 71.30 | 85.00 | 75.10 |
| CWT-SCNN [36] | 74.70 | 81.30 | 68.30 | 96.30 | 92.50 | 86.90 | 73.40 | 91.60 | 84.40 | 83.20 |
| CAgross [37] | 68.31 | 55.10 | 54.61 | 91.14 | 80.17 | 72.23 | 67.79 | 88.65 | 80.23 | 73.13 |
| SCNN [29] | 75.43 | 55.36 | 52.09 | 94.96 | 87.60 | 79.71 | 79.77 | 87.87 | 84.69 | 77.50 |
| SCNN-MCD | 74.84 | 54.34 | 52.44 | 95.84 | 87.32 | 82.04 | 80.42 | 88.34 | 83.92 | 77.72 |