Course details

Classification and recognition

KRD Acad. year 2024/2025 Summer semester

Estimation of parameters Maximum Likelihood and Expectation-Maximization, formulation of the objective function of discriminative training, Maximum Mutual information (MMI) criterion, adaptation of GMM models, transforms of features for recognition, modelling of feature space using discriminative sub-spaces, factor analysis, kernel techniques, calibration and fusion of classifiers, applications in recognition of speech, video and text.

State doctoral exam - topics:

  1. Maximum Likelihood estimation of parameters of a model
  2. Probability distribution from the exponential family and sufficient statistics
  3. Linear regression model and its probabilistic interpretation
  4. Bayesian models considering the probability distribution (uncertainty) of model parameters
  5. Conjugate priors and their significance in Bayesian models
  6. Fishers linear discriminant analysis
  7. Difference between generative and discriminative classifiers; their pros and cons
  8. Perceptron and its learning algorithm as an example of linear classifiers
  9. Generative linear classifier - Gaussian classifier with shared covariance matrix
  10. Discriminative classifier based on linear logistic regression

Guarantor

Burget Lukáš, doc. Ing., Ph.D. (DCGM)

Language of instruction

Czech, English

Completion

Examination

Time span

  • 39 hrs lectures

Assessment points

  • 100 pts final exam

Department

Department of Computer Graphics and Multimedia (UPGM)

Learning objectives

To understand advanced classification and recognition techniques and to learn how to apply the algorithms and methods to problems in speech recognition, computer graphics and natural language processing. To get acquainted with discriminative training and building hybrid systems.
The students will get acquainted with advanced classification and recognition techniques and learn how to apply basic methods in the fields of speech recognition, computer graphics and natural language processing.

The students will learn to solve general problems of classification and recognition.

Prerequisite knowledge and skills

Basic knowledge of statistics, probability theory, mathematical analysis and algebra.

Study literature

  • Ian Goodfellow, Yoshua Bengio, Aaron Courville: Deep Learning, MIT Press, 2016.
  • Simon Haykin: Neural Networks And Learning Machines, Pearson Education; Third edition, 2016.

Syllabus of lectures

  1. Estimation of parameters of Gaussian probability distribution by Maximum Likelihood (ML)
  2. Estimation of parameters of Gaussian Gaussian Mixture Model (GMM) by Expectation-Maximization (EM)
  3. Discriminative training, introduction, formulation of the objective function
  4. Discriminative training with the Maximum Mutual information (MMI) criterion
  5. Adaptation of GMM models- Maximum A-Posteriori (MAP), Maximum Likelihood Linear Regression (MLLR)
  6. Transforms of features for recognition - basis, Principal component analysis (PCA)
  7. Discriminative transforms of features - Linear Discriminant Analysis (LDA) and Heteroscedastic Linear Discriminant Analysis  (HLDA)
  8. Modeling of feature space using discriminative sub-spaces - factor analysis
  9. Kernel techniques, SVM
  10. Calibration and fusion of classifiers
  11. Applications in recognition of speech, video and text
  12. Student presentations I
  13. Student presentations II


Progress assessment

Oral exam.

Course inclusion in study plans

  • Programme DIT, any year of study, Compulsory-Elective group T
  • Programme DIT, any year of study, Compulsory-Elective group T
  • Programme DIT-EN (in English), any year of study, Compulsory-Elective group T
  • Programme DIT-EN (in English), any year of study, Compulsory-Elective group T
  • Programme VTI-DR-4, field DVI4, any year of study, Elective
  • Programme VTI-DR-4, field DVI4, any year of study, Elective
  • Programme VTI-DR-4 (in English), field DVI4, any year of study, Elective
Back to top