Skip to main content
. 2019 Jan 20;19(2):410. doi: 10.3390/s19020410

Table 4.

Description of the CASIA dataset used in our study (unit: image sequences).

CASIA Dataset Training Dataset (20 people) Testing Dataset (30 people) Total
Real Access Presentation Attack Real Access Presentation Attack
Video 60 180 90 270 600
Image Sequence without Data Augmentation 10,940 34,148 16,029 49,694 110,811
Image Sequence with Data Augmentation 65,640 68,296 16,029 49,694 199,659