Table 13.
Comparison of detection error (HTER) produced by our proposed method with the previous study by Peng et al. for the cross-dataset setup (unit: %).
Detection Method | Trained with | Tested with | HTER |
---|---|---|---|
Using LBP + GS-LBP [53] | CASIA | Replay-mobile | 41.25 |
Replay-mobile | CASIA | 48.59 | |
Using LGBP [53] | CASIA | Replay-mobile | 51.29 |
Replay-mobile | CASIA | 50.04 | |
Using CNN [27] | CASIA | Replay-mobile | 21.496 |
Replay-mobile | CASIA | 34.530 | |
Our Proposed Method | CASIA | Replay-mobile | 12.459 |
Replay-mobile | CASIA | 42.785 |