The method in [53] also trains a pose estimator using 300W dataset to as-Dataset LS3D-W is a large-scale 3D face alignment dataset constructed by annotating ...
確定! 回上一頁