@MastersThesis{Fisar-TR-2011-09,
  IS = { zkontrolovano 01 Apr 2012 },
  UPDATE  = { 2011-09-02 },
  author =	 {Fi{\v s}ar, Ond{\v r}ej},
  supervisor =	 {Franc, Vojtech},
  title =	 {Structural classifier for gender recognition},
  school =	 {Center for Machine Perception, K13133 FEE Czech Technical
                  University},
  address =	 {Prague, Czech Republic},
  year =	 {2011},
  month =	 {September},
  day =		 {6},
  type =	 {{MSc Thesis CTU--CMP--2011--09}},
  issn =	 {1213-2365},
  pages =	 {52},
  authorship =	 {100},
  psurl =	 {[Fisar-TR-2011-09.pdf]},
  project =	 {FP7-ICT-247525 HUMAVIPS, PERG04-GA-2008-239455 SEMISOL},
  annote =	 {This thesis deals with the problem of recognizing
   gender of a person from an image of his/her face. One of the major
   problems in the face classification comes from a large image
   variance caused by unknown pose of the recognized face. Existing
   approaches for gender recognition are based on feature classifiers
   trained from examples. The feature classifier are themselves not
   invariant against pose transformations of the input face. There are
   two common strategies to make the classifier invariant. The first
   strategy is based on registering the input images prior to their
   classification. The second strategy is based on generating
   synthetic training examples by applying all pose transformations
   which do not change the class membership. In this thesis we propose
   a new approach based on using a structural classifier which treats
   the unknown face pose as an additional hidden parameter. The
   proposed structural classifier performs face registration and
   classification simultaneously in one step. We formulate learning of
   the parameters of the structural classifier from examples as a
   convex optimization problem. We experimentally compare the proposed
   structural classifier against one baseline approach and one
   state-of-the-art approach on a large corpus of faces. The
   experiments show that proposed structural classifier outperforms
   the state-of-the-art method achieving relative increase of $10\%$
   in the classification accuracy. We also propose an improvement of
   the Bundle Method for Regularized Risk Minimization which is an
   optimization algorithm suitable for solving large instances of the
   learning problem. Preliminary results show that the improved BMRM
   algorithm can significantly reduced the number of iterations of the
   original method. The last result of this thesis is an open source
   library implementing the proposed classifier and its learning
   algorithm. },
  keywords =	 {gender recognition, structural classification},
}