Table 1. Summary of dataset characteristics.
The columns show, respectively: the cardinality constraint; the number of specimens satisfying the constraint; and the rate of the data satisfying the constraint (first row determines the whole set). Nh, Nd and Np are, respectively, the cardinality of pictures from head, dorsum and profile perspective for each specimen.
| Characteristic | # of specimens | Rate |
|---|---|---|
| Nh + Np + Nd ≥ 1 | 44806 | − |
| Nh = Np = Nd = 1 | 35027 | 78% |
| Nh ≥ 1 and Np ≥ 1 and Nd ≥ 1 | 42944 | 96% |
| Nh ≥ 1 | 43970 | 98% |
| Np ≥ 1 | 44372 | 99% |
| Nd ≥ 1 | 43648 | 97% |