Stochastic Processes and Signals

Course description

Introduction:

  • definition of stochastic process and random signal. Introduction of some important issues from mathematical modeling in statistics and probability theory.

Random signals processing:

  • time and sample mean, random signals filtering (Wienner and Kalman filter), probability distribution evaluation (Expectation-Maximization (EM), Maximum A Posteriori (MAP) and »Maximum Likelihood Linear Regression« (MLLR) procedures)

Modeling of stationary and non-stationary stochastic processes:

  • Gauss process, Poisson process, Gauss-Markov process, non-stationary stochastic processes representation using Hidden Markov Models (HMM)

Examples from speech signals processing, modeling of speech perception and production:

source-filter model for speech production, speech perception model and deconvulution of speech signals, time-frequency representations of speech signals, speech detection, speech signal modeling using HMM

Course is carried out on study programme

Objectives and competences

The aim of the course is to recognize and understand advanced methods for representations of stochastic processes and random signals processing with special attention to the examples on speech signals processing.

Learning and teaching methods

Lectures, individual consultations, project work

Intended learning outcomes

Learning outcomes and competences will be knowledge and understanding on:

  • stochastic processes modeling,
  • advanced methods for random signals processing.

Reference nosilca

GRM, Klemen, ŠTRUC, Vitomir. Deep face recognition for surveillance applications. IEEE intelligent systems, ISSN 1541-1672, May/Jun. 2018, vol. 33, iss. 3, str. 46-50.

EMERŠIČ, Žiga, GABRIEL, Luka Lan, ŠTRUC, Vitomir, PEER, Peter. Convolutional encoder-decoder networks for pixel-wise ear detection and segmentation. IET biometrics, ISSN 2047-4946, May 2018, vol. 7, no. 3, str. 175-184.

KOVAČ, Jure, ŠTRUC, Vitomir, PEER, Peter. Frame-based classification for cross-speed gait recognition. Multimedia tools and applications, ISSN 1380-7501, 2017, vol. , no. , str. 1-23

KRAVANJA, Jaka, ŽGANEC, Mario, ŽGANEC GROS, Jerneja, DOBRIŠEK, Simon, ŠTRUC, Vitomir. Exploiting spatio-temporal information for light-plane labeling in depth-image sensors using probabilistic graphical models. Informatica, ISSN 0868-4952, 2016, vol. 27, no. 1, str. 67-84

GOLOB, Žiga, ŽGANEC GROS, Jerneja, ŠTRUC, Vitomir, MIHELIČ, France, DOBRIŠEK, Simon. A composition algorithm of compact finite-state super transducers for grapheme-to-phoneme conversion. V: SOJKA, Petr (ur.), et al. Text, speech and dialogue : proceedings, 19th International Conference, TSD 2016, Brno, Czech Republic, September 12-16, 2016, (Lecture notes in computer science, ISSN 0302-9743, Lecture notes in artificial intelligence, 9924). Switzerland: Springer. cop. 2016, str. 375-382

JUSTIN, Tadej, ŠTRUC, Vitomir, ŽIBERT, Janez, MIHELIČ, France. Development and evaluation of the emotional Slovenian speech database – EmoLUKS. V: KRÁL, Pavel (ur.), MATOUŠEK, Václav (ur.). Text, speech and dialogue : proceedings, 18th International Conference, TSD 2015, Pilsen, Czech Republic, September 14-17, 2015, (Lecture notes in computer science, ISSN 0302-9743, Lecture notes in artificial intelligence, 9302). Cham [etc.]: Springer. cop. 2015, str. 351-359

Study materials

Robert MG, Lee DD (2004) An Introduction to Statistical Signal Processing. Cambridge University Press

Shlomo E (2007) Random signals and noise: a mathematical introduction. CRC Press

Rabiner L, Schafer R (2010) Theory and Applications of Digital Speech Processing. Prentince Hall

Pieraccini R (2012) The Voice in the Machine: Building Computers That Understand Speech. MIT Press

Bodi na tekočem

Univerza v Ljubljani, Fakulteta za elektrotehniko, Tržaška cesta 25, 1000 Ljubljana

E:  dekanat@fe.uni-lj.si T:  01 4768 411