Robust face alignment and tracking by combining local search and global fitting
作者：
Highlights：
• We propose a new face alignment method that combines local search and global fitting.
• Local misalignments in the local search are restricted by holistic appearance fitting.
• Divergent alignments in the global fitting are avoided by the restricting local movements.
• We extend the proposed face alignment to face tracking by adding a template image.
• The proposed method yields accurate and stable alignment under heavy occlusion.
摘要
•We propose a new face alignment method that combines local search and global fitting.•Local misalignments in the local search are restricted by holistic appearance fitting.•Divergent alignments in the global fitting are avoided by the restricting local movements.•We extend the proposed face alignment to face tracking by adding a template image.•The proposed method yields accurate and stable alignment under heavy occlusion.
论文关键词：Face alignment,Face tracking,Local search method,Global fitting method,Combination of local search and global fitting methods
论文评审过程：Received 22 May 2015, Revised 7 April 2016, Accepted 15 April 2016, Available online 3 May 2016, Version of Record 13 May 2016.
论文官网地址：https://doi.org/10.1016/j.imavis.2016.04.012
