Mathematical Image Analysis Group, Saarland University

Welcome to the homepage of the lecture

Numerical Algorithms for Visual Computing III: Optimisation

Summer Term 2011

Numerical Algorithms for Visual Computing III: Optimisation

Dr. Michael Breuß and M.Sc. Kai Hagenburg

Summer 2011

Lectures (2h) with exercises (2h), summer term 2011.

Lectures: Th 12–14 (12 p.m.–14 p.m.), Building E2.4, Lecture Hall IV
Tutorial: Mo 14–16 (14 p.m.–16 p.m.), Building E1.3, Seminar Room 015

The final results after the second written examination are now available.
You can find a list with the results here.

Course in numerical methods useful in the field of visual computing. The third part of the course is concerned with optimisation methods.

Entrance requirements – Contents – Assessments / Exams – References – Downloads

Entrance requirements

Undergraduate knowledge of mathematics. For computer science students, this requirement is met by having completed the Mathematics for Computer Scientists lecture cycle.

Mathematical prerequisites which exceed the basic mathematics courses are provided within the lecture. Previous knowledge is not required.

Contents

The course is concerned with numerical methods useful for image analysis (image processing, computer vision, pattern recognition) and synthesis (computer graphics, scientific visualisation, geometric modelling).

Topics include:

variational optimisation methods
discrete optimisation methods

Assessments / Exams

There will be two written exams, the first one at the end of the semester instead of the last lecture. The date of the second written exam will be announced. The second written examination will take place on September 22, starting at 14:15 in the seminar room 3.23 in the MMCI cluster building E1.7 (3rd floor). Please make sure that you are there in time, ideally at 14:00.
Concerning the allowed material, you may bring along the course material from the lectures and tutorials, nothing else.
Please note that for NAVC3, we expect our students to achieve fifty percent of the total points one may obtain by submitting results of exercises. This can be done in groups of up to three students.

References

Diverse books and scripts on numerical mathematics, depending on the topic. Details will be given in the lecture.

Downloads

The script is available here (Current version: July 12th). It will be updated regularly in the course of the semester. Of course, we do not take on any responsibilities for (rare) typos.

The SciLab tutorial is available here. Of course, we do not take on any responsibilities for (rare) typos. Supplementary, the SciLab-file used in the tutorial for illustration can be accessed here.

Participants of the course can download the assignment sheets here:

No. Title Deadline Solution Code

1 Assignment 1 April 23 Solution ---

2 Assignment 2 May 12 Solution ---

3 Assignment 3
Programming files May 26 Solution Code

4 Assignment 4 June 09 Solution Code

5 Assignment 5 June 16 Solution

5 Assignment 6 July 1 Solution

6 Assignment 7 July 14 Solution

7 Assignment 8 (updated) July 18 Solution

In case of questions to SCILAB, you may readily contact Kai Hagenburg under hagenburg -at- mia.uni-saarland.de (please replace anti-spam at) .