声明
Acknowledgements
Abstract
Contents
List of Figures
List of Tbles
Nomenclature
Chapter 1 Introduction
1.1 Introduction
1.2 Pattern recognition process
1.3 Problem statement
1.4 Thesis contributions and outline
Chapter 2 Supervised vs.unsupervised learning
2.1 Introduction
2.2 Supervised algorithms
2.2.1 Decision tree learner
2.2.2 Adaboost
2.2.3 Support Vector Machine
2.2.4 Artificial Neural Networks
2.2.5 k-Nearest Neighbor
2.3 Unsupervised algorithms
2.3.1 Hierarchical Clustering
2.3.2 k-means
2.3.3 Self-Organizing Maps
Chapter 3 Generalizing binary classifiers to multiclass classifiers
3.1 Introduction
3.2 Decomposition strategies
3.2.1 One-Against-All
3.2.2 One-Against-One
3.2.3 Error-Correcting Output Code
3.3 Hierarchical strategies
3.4 Stacked Generalization
3.5 Single optimization strategy
Chapter 4 MPC-based multiclass classification
4.1 Meta Probability Code
4.2 MPC-based algorithm
4.2.1 Overview
4.2.2 Cluster post-processing
4.2.3 Toy example
4.3 Experiments and Results
4.3.1 Datasets
4.3.2 Effectiveness of the proposed approach
4.3.3 Classification results
4.3.4 Comparison with other multi-class classifiers
4.4 Conclusion
Chapter 5 Application to facial expression recognition
5.1 Introduction
5.2 Literature review
5.3 MPC-based FER framework
5.4 Features for facial representation
5.4.1 Local Binary Pattern
5.4.2 Gabor-wavelet
5.4.3 Zernike moments
5.4.4 Facial fiducial points
5.5 Experimental studies
5.5.1 Datasets
5.5.2 Performance evaluation
5.5.3 Statistical comparison of the FER systems
5.5.4 Generalization performance on across datasets
5.6 Conclusion
Chapter 6 Application to face recognition
6.1 Introduction
6.2 Literature review
6.3 Experimental studies
6.3.1 Datasets
6.3.2 Performance evaluation
6.4 Conclusion
Chapter 7 Conclusion and future direction
7.1 Conclusion
7.2 Future work
Bibliography
List of publications