Taught Post Graduate Courses

Introduction to Statistical models.
Bayes decision theory, maximum likelihood estimation, Bayes estimation, expectation maximization algorithm, hidden Markov models.
Linear Classifiers, feature selection and reduction with transforms, Principal Component Analysis (PCA).
Unsupervised learning, clustering and non-parametric classifiers.
Clustering models with k-means and Nearest Neighbor.
Linear classifiers, Perceptron algorithm and support vector machines (SVM).
Linear regression.
Non-linear classifiers and Artificial Neural Networks (ANNs).
Deep Neural Network models.
Non-metric methods, classification and regression trees.
Bayesian networks, non-parametric methods and Parzen windows.
Accuracy estimation, cross-validation and Receiver Operator Characteristic curves (ROC curves).

Principles and methodological concepts in machine vision with emphasis on algorithms and applications.
Image formation and models (geometric, color, frequency, symbolic).
Basic image processing methods including filtering, normalization, enhancement, edge detection with first and second derivative operators, image thresholding and content enhancement.
Image segmentation and edge models using split and merge, hierarchical segmentation, relaxation labeling, Hough transform.
Binary image processing, distance and morphological transforms, shape recognition and region labeling.
Image representation and understanding.
Color and texture analysis for content representation and modeling.
Texture understanding with structural and statistical methods.
Dynamic vision effects, motion estimation, optical flow and motion tracking.
Principles of video analysis and applications in information systems.
3D projections from photometric stereo and motion.
Recovering shape, orientation and motion of 3D objects, with applications in robotics and automation.