Neural network ensembles for image identification using Pareto-optimal features

In this paper, an ensemble classifier is constructed for invariant image identification, where the inputs to the ensemble members are a set of Pareto-optimal image features extracted by an evolutionary multi-objective Trace transform algorithm. The Pareto-optimal feature set, called Triple features, gains various degrees of trade-off between sensitivity and invariance. Multilayer perceptron neural networks are adopted as ensemble members due to their simplicity and capability for pattern classification. The diversity of the ensemble is mainly achieved by the Pareto-optimal features extracted by the multi-objective evolutionary Trace transform. Empirical results show that the general performance of proposed ensemble classifiers is more robust to geometric deformations and noise in images compared to single neural network classifiers using one image feature.

[1]  Wenjia Wang,et al.  Investigation on Diversity in Homogeneous and Heterogeneous Ensembles , 2006, The 2006 IEEE International Joint Conference on Neural Network Proceedings.

[2]  Xin Yao,et al.  Ieee Transactions on Knowledge and Data Engineering 1 Relationships between Diversity of Classification Ensembles and Single-class Performance Measures , 2022 .

[3]  Ashish Ghosh,et al.  Ensemble of Multilayer Perceptrons for Change Detection in Remotely Sensed Images , 2014, IEEE Geoscience and Remote Sensing Letters.

[4]  Kin Keung Lai,et al.  Multistage RBF neural network ensemble learning for exchange rates forecasting , 2008, Neurocomputing.

[5]  Ludmila I. Kuncheva,et al.  Measures of Diversity in Classifier Ensembles and Their Relationship with the Ensemble Accuracy , 2003, Machine Learning.

[6]  Michal Wozniak,et al.  A framework for image analysis and object recognition in industrial applications with the ensemble of classifiers , 2013, 2013 IEEE 18th Conference on Emerging Technologies & Factory Automation (ETFA).

[7]  Mark Johnston,et al.  Evolving Diverse Ensembles Using Genetic Programming for Classification With Unbalanced Data , 2013, IEEE Transactions on Evolutionary Computation.

[8]  Bernhard Sendhoff,et al.  Pareto-Based Multiobjective Machine Learning: An Overview and Case Studies , 2008, IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews).

[9]  Alexander Kadyrov,et al.  The Trace Transform and Its Applications , 2001, IEEE Trans. Pattern Anal. Mach. Intell..

[10]  Jessica J. Fridrich,et al.  Ensemble Classifiers for Steganalysis of Digital Media , 2012, IEEE Transactions on Information Forensics and Security.

[11]  Bernhard Sendhoff,et al.  Corrections to “Pareto-Based Multiobjective Machine Learning: An Overview and Case Studies” [May 08 397-415] , 2009, IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews).

[12]  Kalyanmoy Deb,et al.  A fast and elitist multiobjective genetic algorithm: NSGA-II , 2002, IEEE Trans. Evol. Comput..

[13]  Tony R. Martinez,et al.  Decision Tree Ensemble: Small Heterogeneous Is Better Than Large Homogeneous , 2008, 2008 Seventh International Conference on Machine Learning and Applications.

[14]  Lars Kai Hansen,et al.  Neural Network Ensembles , 1990, IEEE Trans. Pattern Anal. Mach. Intell..

[15]  Min Liu,et al.  The RST invariant digital image watermarking using Radon transforms and complex moments , 2010, Digit. Signal Process..

[16]  Yaochu Jin,et al.  Evolutionary multi-objective optimization of trace transform for invariant feature extraction , 2012, 2012 IEEE Congress on Evolutionary Computation.

[17]  Jian Wang,et al.  Number Image Recognition Based on Neural Network Ensemble , 2007, Third International Conference on Natural Computation (ICNC 2007).

[18]  Yaochu Jin,et al.  A Comparative Study of Multi-objective Evolutionary Trace Transform Methods for Robust Feature Extraction , 2013, EMO.

[19]  P. Toft.  The Radon Transform - Theory and Implementation , 1996 .

[20]  Yaochu Jin,et al.  Evolutionary Multiobjective Image Feature Extraction in the Presence of Noise , 2015, IEEE Transactions on Cybernetics.

[21]  Aleksandra Pizurica,et al.  Object identification by using orthonormal circus functions from the trace transform , 2012, 2012 19th IEEE International Conference on Image Processing.

[22]  Hong Pan,et al.  Efficient Object Recognition Using Boundary Representation and Wavelet Neural Network , 2008, IEEE Transactions on Neural Networks.

[23]  Xin Yao,et al.  Making use of population information in evolutionary artificial neural networks , 1998, IEEE Trans. Syst. Man Cybern. Part B.

[24]  David W. Aha,et al.  Ensembles of Classifiers for Morphological Galaxy Classification , 2001 .