Table 7.
Description of subsets of the CASIA dataset used in our study (unit: image sequences).
| Dataset Name | Training Dataset (20 people) | Testing Dataset (30 people) | Total | ||
|---|---|---|---|---|---|
| Real Access | Presentation Attack | Real Access | Presentation Attack | ||
| Low Quality Dataset | 3140 | 11,019 | 5298 | 16,166 | 35,623 |
| Normal Quality Dataset | 3223 | 11,275 | 4949 | 16,141 | 35,588 |
| High Quality Dataset | 4577 | 11,854 | 5782 | 17,387 | 39,600 |
| Wrap-Photo Dataset | 10,940 | 12,871 | 16,029 | 19,271 | 59,111 |
| Cut-Photo Dataset | 10,940 | 9499 | 16,029 | 14,784 | 51,252 |
| Video Display Dataset | 10,940 | 11,778 | 16,029 | 15,639 | 54,386 |