We present a novel approach for automatically detecting and tracking facial landmarks across poses and expressions from in-the-wild monocular video data, e.g.,
Deformation of skin and muscle is essential for bringing an animated character to life. This deformation is difficult to animate in a realistic fashion using traditional techniques because of the subt