Joint subspace learning for reconstruction of 3D facial dynamic expression from single image | IEEE Conference Publication | IEEE Xplore