Newton Optimization Based Congealing for Facial Image Alignment - ACROE - Ingénierie de la Création Artistique Accéder directement au contenu
Communication Dans Un Congrès Année : 2011

Newton Optimization Based Congealing for Facial Image Alignment

Weiyuan Ni
  • Fonction : Auteur
  • PersonId : 908936

Résumé

Congealing is an unsupervised image alignment method for a set of images, and the transformation parameters are obtained by minimizing a sum-of-entropies function. In this paper, we provide a solution to improve the estimation of transformation parameters using a Newton optimization method, under the premise of maintaining the compatibility with feature descriptors. Besides, instead of SIFT descriptor in canonical Congealing, we combine Congealing with POEM (Patterns of Oriented Edge Magnitudes) which catches both edge information and the relation between pixels at a neighboring region. The experiment results show that our alignment method has better ability for the removal of unwanted displacements, and also improves the performance of face recognition.
Fichier non déposé

Dates et versions

hal-00617619 , version 1 (29-08-2011)

Identifiants

Citer

Weiyuan Ni, Alice Caplier. Newton Optimization Based Congealing for Facial Image Alignment. ICIP 2011 - 18th IEEE International Conference on Image Processing, Sep 2011, Bruxelles, Belgium. pp.577-580, ⟨10.1109/ICIP.2011.6116614⟩. ⟨hal-00617619⟩
145 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More