This paper describes a robust algorithm for automatically learning an appearance subspace of objects performing rigid motion through an image sequence, given a manual initialization of the regions of support (masks) in the first frame. The learning process is posed as a continuous optimization problem and it is solved with a mixture of stochastic and deterministic techniques achieving sub-pixel accuracy. Additionally, we learn the dynamics of the motion and appearance parameters for scene characterization and point out the benefits of working with modular eigenspaces. Preliminary results of automatic learning a modular eigenface model with applications to real time video conferencing, human computer interaction and actor animation are reported.
Automatic learning of appearance face models
2001-01-01
1186077 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Automatic Learning of Appearance Face Models
British Library Conference Proceedings | 2001
|Automatic face identification system using flexible appearance models
British Library Online Contents | 1995
|Generative face alignment through 2.5D active appearance models
British Library Online Contents | 2013
|Towards a low bandwidth talking face using appearance models
British Library Online Contents | 2003
|Learning appearance models for road detection
IEEE | 2013
|