A PSO Based Adaboost Approach to Object Detection

This paper describes a new approach using particle swarm optimisation (PSO) within AdaBoost for object detection. Instead of using the time consuming exhaustive search for finding good features to be used for constructing weak classifiers in AdaBoost, we propose two PSO based methods in this paper. The first uses PSO to evolve and select the good features only and the weak classifiers use a kind of decision stump. The second uses PSO for both selecting the good features and evolving weak classifiers in parallel. These two methods are examined and compared on a pasta detection data set. The experiment results show that both approaches perform quite well for the pasta detection problem, and that using PSO for selecting good individual features and evolving associated weak classifiers in AdaBoost is more effective than for selecting features only for this problem.

[1]  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.

[2]  D. Opitz,et al.  Popular Ensemble Methods: An Empirical Study , 1999, J. Artif. Intell. Res..

[3]  Giorgio Valentini,et al.  Ensembles of Learning Machines , 2002, WIRN.

[4]  Anders Krogh,et al.  Neural Network Ensembles, Cross Validation, and Active Learning , 1994, NIPS.

[5]  Alejandro Echeverría,et al.  Evolutionary discriminant analysis , 2006, IEEE Transactions on Evolutionary Computation.

[6]  Robert E. Schapire,et al.  The Boosting Approach to Machine Learning An Overview , 2003 .

[7]  Leslie G. Valiant,et al.  Cryptographic limitations on learning Boolean formulae and finite automata , 1994, JACM.

[8]  John Plaice,et al.  Distributed Communities on the Web , 2002, Lecture Notes in Computer Science.

[9]  Chuanyi Ji,et al.  Combinations of Weak Classifiers , 1996, NIPS.

[10]  Stefano Cagnoni,et al.  Particle Swarm Optimization for Object Detection and Segmentation , 2007, EvoWorkshops.

[11]  Leslie G. Valiant,et al.  A theory of the learnable , 1984, STOC '84.

[12]  Gary Bradski,et al.  Learning-Based Computer Vision with Intels Open Source Computer Vision Library , 2005 .

[13]  Rolf Drechsler,et al.  Applications of Evolutionary Computing, EvoWorkshops 2008: EvoCOMNET, EvoFIN, EvoHOT, EvoIASP, EvoMUSART, EvoNUM, EvoSTOC, and EvoTransLog, Naples, Italy, March 26-28, 2008. Proceedings , 2008, EvoWorkshops.

[14]  David Bargeron,et al.  Boosting-based transductive learning for text detection , 2005, Eighth International Conference on Document Analysis and Recognition (ICDAR'05).

[15]  Yoav Freund,et al.  A decision-theoretic generalization of on-line learning and an application to boosting , 1997, EuroCOLT.

[16]  Xin Yao,et al.  Diversity creation methods: a survey and categorisation , 2004, Inf. Fusion.

[17]  James Kennedy,et al.  Particle swarm optimization , 2002, Proceedings of ICNN'95 - International Conference on Neural Networks.

[18]  Leo Breiman,et al.  Bagging Predictors , 1996, Machine Learning.

[19]  David D. Denison,et al.  Nonlinear estimation and classification , 2003 .