@MastersThesis{Fajt-CAK-2007-27,
  author =       {Fajt, Luk{\'a}{\v s}},
  supervisor =   {Svoboda, Tom{\'a}{\v s}},
  title =        {Pictorial Structural Models, Learning and
                  Recognition in Image Sequences},
  school =       {Department of Cybernetics, Faculty of Electrical,
                  Engineering Czech Technical University},
  address =      {Prague, Czech Republic},
  year =         {2007},
  month =        {February},
  day =          {15},
  type =         {{MSc Thesis K333--27/07, CTU--CMP--2007--04}},
  pages =        {65},
  authorship =   {100},
  psurl =        {[Fajt-TR-2007-04.pdf]},
  project =      {1M0567, 1ET101210407},
  annote = {This paper describes the detector of the people as
    assemblies of individual parts as head, torso, limbs, etc. We
    built on Felzenszwalb and Huttenlocher's approach for efficient
    assembling of candidate parts into pictorial structures. A human
    body is represented as a collection of parts arranged in a
    deformable configuration. The appearance of each part is modeled
    separately, and the deformable configuration is represented by
    spring-like connections between pairs of parts. We implemented
    three different appearance part detectors. First detector is
    simple, it requires segmented data - it is obtained with
    background subtraction. Second one uses more flexible colour based
    segmentation of parts. The third one uses SVM classifier for
    computing probability, important features are extracted from
    filtered picture as brightness values of pixels. Probability model
    of appearance of parts and joints between pairs of parts is
    learned from manually labelled training data. The best match is
    chosen in two ways, as the one maximising a posterior probability
    or the 100 samples from posterior probability are taken with Monte
    Carlo method. Then the best match is selected by another
    method. We propose a new method for selecting the best match from
    samples based on epipolar geometry, two calibrated cameras are
    needed. Distance of significant points of the model from
    corresponding epipolar line is measured. },
  keywords = {computer vision, human-computer interfaces, 
              pictorial model, articulated model},
}