Table 4 Performance comparison of multiple models on validation set.

From: Middle-shallow feature aggregation in multimodality for face anti-spoofing

Model

ACER

TPR@FPR = 10E−2

TPR@FPR = 10E−3

TPR@FPR = 10E−4

MSFA

0.00079

1.0000

0.9997

0.9990

Baseline

0.00158

0.9999

0.9976

0.9884

MobilenetV2

0.01245

0.9850

0.94367

0.85467

FishNet

0.02362

0.9610

0.9010

0.7490