作者: Moritz Kaiser , Dejan Arsic , Shamik Sural , Gerhard Rigoll
DOI: 10.1109/ICIP.2010.5654057
关键词:
摘要: Exact 3D tracking of facial feature points is appealing for many applications in human-machine interaction. In this work a Active Shape Model (ASM) that can be shifted, scaled, and rotated used to track the points. The efficient Gauss-Newton method applied estimate ASM, rotation, translation, scale parameters. If head turns one side, some might occluded but they are still considered estimation A robust error norm reduces (or ideally cancels) influence applied. With algebraic transformations computational cost per frame further reduced. proposed algorithm evaluated on basis Airplane Behavior Corpus.