However, their dataset was composed of relatively few images (dozens of individuals) and not obtained from a real scenario. In another work on cross-domain face ...
確定! 回上一頁