This seminar discusses individual algorithms and applications of computer vision. The goal of this seminar is to introduce students to selected aspects of computer vision. In this term, we will cover algorithms for 2.5D scene analysis, forgery detection for video sequences and potentially other current research topics.

Learning Objectives

• Perform literature research based on a given scientific article
• Independently study the topic based on the found literature
• Develop a good understanding of the algorithm/method and potentially implement it
• Present the given topic in a manner that is understandable for their peers
• Give a talk in English and gain language competence

Requirements

Experience with C++ or C# is expected. Basic knowledge of image processing is desirable. Having visited lectures such as Introduction to Pattern Recognition (IntroPR) or Diagnostic Medical Image Processing (DMIP) might prove beneficial.

Please contact us if for additional information: Fürsattel, Peter or Mullan, Patrick.

Organization 

In the beginning of the term, there will be an introductory meeting to discuss the organization of the project and for presenting the individual topics.

During the term you will have time to prepare a presentation and your implementation/paper. The final presentations will be held at the end of the term.

The grade consists of two parts: the final presentation of the topic and a seminar paper or an implementation. Each participant can choose between writing a 6 pages seminar paper (IEEE style), or creating an implementation/demonstrator, including a short, informal documentation.

Registration

It is required to register for the seminar via StudOn. Registration will open on Sep. 15th 12:00.

First Meeting

The first meeting will be on Oct., 24th, 12:00 (s.t.) to 14:00, in room 01.134. This information is subject to change, please check back in October.