Because variations of face images show a complicated nonlinear manifold in high-dimensional data space, we use an LLE (Locally Linear Embedding) technique ...
確定! 回上一頁