Real-time combined 2d+3d active appearance models. (2004)

by J Xiao, S Baker, I Matthews, T Kanade
Venue:In IEEE computer vision and pattern recognition (Vol. II,