Table 5 State of the art methods which assume a client/subject independent frame and the corresponding error rates
From: Image life trails based on contrast reduction models for face counter-spoofing
Method | Classifier | Train data | Test data | Threshold | EER |
|---|---|---|---|---|---|
LBP [36] | GMMUBM | CASIA | CASIA | \(C_{gb}\) | 21.69 |
Two-class SVM | CASIA | CASIA | \(C_{gb}\) | 15.42 | |
LBP-TOP [36] | GMM UBM | CASIA | CASIA | \(C_{gb}\) | 12.65 |
Two-class SVM | CASIA | CASIA | \(C_{gb}\) | 8.53 | |
Motion [36] | GMMUBM | CASIA | CASIA | \(C_{gb}\) | 12.52 |
Two-class SVM | CASIA | CASIA | \(C_{gb}\) | 11.53 | |
IMQ [11] | One-class SVM | CASIA | CASIA | - | 23.07 |
BSIF [11] | One-class SVM | CASIA | CASIA | - | 36.06 |
LPQ [11] | One-class SVM | CASIA | CASIA | - | 35.19 |
LBP [11] | One-clas SVM | CASIA | CASIA | - | 25.06 |
Random scan [4] | One-class SVM | CASIA | CASIA | \(N_{S}=1\) | 3.5122 |
Random scan [4] | One-class SVM | CASIA | CASIA | \(N_{S}=20\) | 1.8920 |