Computer Vision and Pattern Recognition

An Advanced Course covering the depth of principles and applications of Pattern Recognition


Course Code:              EE7107
Credit Hours:                       3

Probability and Stochastic Processes

Target Audience:
MS/PhD students wishing to pursue research in the field of Pattern Recognition.


In a number of applied fields of electrical and computer engineering, there are problems where a decision has to be made with limited knowledge.  These are typically pattern recognition problems and can adequately be formulated in the context of a decision theory.  The course “Computer Vision and Pattern Recognition” aims to develop the analytical skills of taking the best decision based on certain cost function(s).  The course therefore starts with an intensive description of Bayesian Decision Theory followed by statistical techniques of data analysis such as principal component analysis, linear discriminant analysis etc.  Parameter estimation with the knowledge of prior probabilities is further discussed.  Problems where assumption of distribution over a given dataset is not possible are dealt using the concept of parametric estimation.  Applications of these algorithms in applied fields such as face recognition, speaker identification, texture classification and multiscale analysis are covered to highlight the pragmatic relevance of the course.  Special emphasis on course project is given and students will be guided to publish in reputable journals and conferences.


The course will be taught by Dr Imran Naseem (  He holds a PhD degree in Electrical, Electronics and Computer Engineering from the University of Western Australia (UWA).  Dr Imran has also worked as a Research Fellow at Curtin University of Technology, Australia.  He has published in the highest impact factor journals such as IEEE Transactions on Pattern Analysis and Machine Intelligence and Pattern Recognition Journal.  He has a number of research papers and book chapters to his credit including the prestigious IEEE ICIP conference.  Current focus of his research is on multimedia signal processing including face recognition and texture analysis.

Course Outline:

  • Classification theory in terms of Bayesian costs, decision functions and the geometry of decision regions for continuous and discrete random variables, Classification error probabilities and bounds
  • Maximum-likelihood and Bayesian parameter estimation
  • Nonparametric recognition, Parzen window operation, K-nearest neighbor classifier
  • Decision Trees
  • Algorithm independent machine learning, resampling for estimating statistics and accuracy
  • Mixture densities and identifiability, K-Means clustering, unsupervised Bayesian learning, decision-directed approximation, hierarchical clustering, minimum spanning trees.
  • Applications to computer vision problems of estimation and recognition



Recommended Books:

  • Pattern Classification, 2nd edition, by R. Duda, P. Hart, D. Stork.
  • Pattern Recognition and Machine Learning, by C. M. Bishop.