Datorseende (CVL)

Objectives

This course gives a comprehensive introduction to detection (decision-making) as well as parameter estimation and signal estimation (filtering) based on observations of discrete-time and continuous-time signals. The material taught in this course has applications in many areas, for example in communications, radar, and in control, but also in pattern recognition and imaging.

Instructor

Prof. Rudolf Mester , ISY/ Computer Vision Laboratory

Registration

The course is open to students enrolled in a Ph.D. program at Linkoping University/ISY. External participants are welcome in accordance to the usual LIU practice.

If you have interest to participate, please register by sending an email to the instructor, by Monday, Aug 12, 2013. Register as early as possible in order to be informed about any updates/changes via email.

Prerequisites

• Good knowledge of linear algebra, probability, and stochastic processes. General mathematical maturity.

Course outline (tentative)

• Binary and M-ary hypothesis testing
• Detection theory: Neyman-Pearson, ROC, Bayesian criteria
• Estimation theory: classical estimation, maximum likelihood, Cramer-Rao lower bound, Bayesian estimation, MMSE
• Composite hypothesis testing, model order selection
• General Gaussian models
• Representation of continuous-time waveforms and noise (Karhunen-Loève expansion)
• Detection and parameter estimation of signals in additive Gaussian noise
• Estimation of continuous-time and discrete-time random processes (Gauss-Markov processes, Wiener and Kalman filters)
  

Schedule

The course consists of 12 seminars / meetings between end of August 2013 and November 2013. Planned course start is Wednesday August 28, 2013 (time and place to be specified soon)

A course schedule (with reading and homework) can be found here soon (this information will be continuously updated during the course).

Literature

• H. L. van Trees, K. Bell, Z. Tian: "Detection, estimation and modulation theory: Part I", Wiley (2nd edition ISBN 978-0-470-54296-5, 2013.)
• Course notes (will be continuously updated during the course) are here (soon)
  
• Additional homework problems are here (soon)
  
• Supplementary material handed out during the course and/or available from this webpage

Credits

The course is estimated to be worth 12 ECTS credits.

Examination

• 6 ECTS version:
• Active seminar participation.
• Solve 40% of the problems for each homework assignment.
• Oral exam, when necessary.

• 12 ECTS version:
• Active seminar participation
• Solve 80% of the problems for each homework assignment.
• Present one or more problem per homework seminar on the board.
• Oral exam, when necessary.

Homework to be solved, written up and submitted individually. Deadlines to be strictly observed.