Face registration

Object alignment methods aim at finding the transformation or deformation which minimizes the discrepancies between two or more objects. In automated face analysis, these discrepancies usually stem from rigid head motions induced by observing faces at different time instances and/or from different viewpoints  as well as from non-rigid facial deformations typically induced by facial muscle contractions. Alignment methods are the key components for estimating these motions and, therefore, play a central role in the efficacy of high-level applications such as face recognition, audio-visual speech recognition and analysis of human behavior. The group's research focuses on parametric object alignment methods particularly suitable for face analysis applications which require efficiency and robustness against occlusions and illumination changes.

Involved group members

Maja Pantic, Joan Alabort-i-Medina, Epameinondas Antonakos, Akshay Asthana, James Booth, Shiyang Cheng, Ioannis Marras, Christos Sagonas, Georgios Tzimiropoulos

Related Publications

1. Robust and Efficient Parametric Face Alignment

   G. Tzimiropoulos, S. Zafeiriou, M. Pantic. Proceedings of IEEE Int’l Conf. on Computer Vision (ICCV 2011) . pp. 1847 - 1854, Oral. November 2011.
   [pdf] [bibtex] [endnote]
   Bibtex reference [hide]

   @inproceedings{tzimiroREPFAICCV2011,
       author = {G. Tzimiropoulos and S. Zafeiriou and M. Pantic},
       pages = {1847--1854},
       booktitle = {Proceedings of IEEE Int’l Conf. on Computer Vision (ICCV 2011) },
       month = {November},
       note = {Oral},
       title = {Robust and Efficient Parametric Face Alignment},
       year = {2011},
   }
   Endnote reference [hide]

   %0 Conference Proceedings
   %T Robust and Efficient Parametric Face Alignment
   %A Tzimiropoulos, G.
   %A Zafeiriou, S.
   %A Pantic, M.
   %B Proceedings of IEEE Int’l Conf. on Computer Vision (ICCV 2011)
   %D 2011
   %8 November
   %F tzimiroREPFAICCV2011
   %O Oral
   %P 1847-1854