Probabilistic Darwin Machines for Object Detection

Since the apparition of the first object detection systems in the late 1960s, the complexity of faced scenarios has been continuously growing. From a small set of simple objects in a homogeneous background, where simple alignment strategies or the use of geometric primitives were enough to recognize them, to nowadays, where we are dealing with sets of thousands of complex objects in cluttered scenes, where more sophisticated methods to describe and recognize objects are necessary. In this paper, we present an approach to the object detection problem based on Probabilistic Darwin Machines (PDM), where one of the most used object detection systems has been redefined in order to allow the use large feature sets, which are able to better describe the objects. Two different PDM are used in combination with a large set of visual features in order to learn an object detection system for five different real world visual object classes.

[1]  Shumeet Baluja,et al.  Using Optimal Dependency-Trees for Combinational Optimization , 1997, ICML.

[2]  David E. Goldberg,et al.  The compact genetic algorithm , 1999, IEEE Trans. Evol. Comput..

[3]  H. Muhlenbein,et al.  The Factorized Distribution Algorithm for additively decomposed functions , 1999, Proceedings of the 1999 Congress on Evolutionary Computation-CEC99 (Cat. No. 99TH8406).

[4]  W. Calvin.  The Six Essentials? Minimal Requirements for the Darwinian Bootstrapping of Quality , 1997 .

[5]  Rainer Lienhart,et al.  An extended set of Haar-like features for rapid object detection , 2002, Proceedings. International Conference on Image Processing.

[6]  Shimon Ullman,et al.  Object Classification Using a Fragment-Based Representation , 2000, Biologically Motivated Computer Vision.

[7]  M. Pelikán,et al.  The Bivariate Marginal Distribution Algorithm , 1999 .

[8]  Janez Demsar,et al.  Statistical Comparisons of Classifiers over Multiple Data Sets , 2006, J. Mach. Learn. Res..

[9]  Rich Caruana,et al.  Removing the Genetics from the Standard Genetic Algorithm , 1995, ICML.

[10]  Y. Freund,et al.  Discussion of the Paper \additive Logistic Regression: a Statistical View of Boosting" By , 2000 .

[11]  G. Harik.  Linkage Learning via Probabilistic Modeling in the ECGA , 1999 .

[12]  Paul A. Viola,et al.  Rapid object detection using a boosted cascade of simple features , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[13]  Heinz Mühlenbein,et al.  The Equation for Response to Selection and Its Use for Prediction , 1997, Evolutionary Computation.

[14]  Paul A. Viola,et al.  MIMIC: Finding Optima by Estimating Probability Densities , 1996, NIPS.

[15]  Pawan Sinha,et al.  Qualitative Representations for Recognition , 2002, Biologically Motivated Computer Vision.

[16]  Yoav Freund,et al.  Experiments with a New Boosting Algorithm , 1996, ICML.

[17]  David G. Lowe,et al.  Object recognition from local scale-invariant features , 1999, Proceedings of the Seventh IEEE International Conference on Computer Vision.

[18]  S. Blackmore.  The Meme Machine , 1999 .

[19]  P. Sinha,et al.  Dissociated Dipoles: Image Representation via Non-local Comparisons , 2003 .

[20]  Pedro M. Domingos,et al.  Naive Bayes models for probability estimation , 2005, ICML.

[21]  D. Goldberg,et al.  BOA: the Bayesian optimization algorithm , 1999 .