186.117 Forschungsseminar aus Computergraphik und digitale Bildverarbeitung WS+SS 2.0h
Research Seminar on Computer Graphics and Image Processing
HOT NEWS
>>> Initial meeting (Vorbesprechung) <<<
on Thursday, 03.03.2011 at 5 PM (17:00 Uhr),
in the seminar room of the Institute 186 (Favoritenstr. 9-11/5.floor).
In this initial meeting we will define working groups and present/develop a topic for each group.
Here are the power point slides of the initial meeting.
Contact: Stefan Jeschke.
General information
The goal of this seminar is to teach you scientific communication skills and basic scientific work flow. In principle, you work on a very specific topic covered by one or very few research papers. Each topic is supervised by a researcher of our institute, who will be your advisor. There might be related topics for different students (up to 3). There will be several meetings with your advisor and other students during the semester where you have to present your ideas/progress/problems/etc.. You will also prepare a final presentation of your work in the form of a talk or a written scientific article.
- Grundlagen methodischen Arbeitens.
- Basic knowledge in computer graphics (according to our course Computergraphik 1.)
Procedure:
- Agree on a topic with any researcher at the institute.
- Attend to the following talks:
- Friday, 11.03. at 14:15: "Forschung und wie sie funktioniert" (Prof.Gröller) in EI2
- Friday, 15.04. at 14:15: "Wie schreibt man eine wissenschaftliche Arbeit?" (Prof. Breiteneder), in EI2
- Monday, 16.05. at 14:00: "Wie halte ich einen Vortrag?" (Prof.Purgathofer) in HS7
- Everyone will try to analyse one or two recent scientific paper(s), to be selected with the advisor.
- Study of literature relevant for understanding the paper (if necessary).
- Evaluate the paper results: either compare with existing results / code from the web / implement it (if this is a reasonable amount of work).
- Things to analyse during an evaluation:
- Does the algorithm work as described?
- How do parameters influence the result?
- For which 3D models/images/constraints/etc does the proposed algorithm fail?
- How fast is the algorithm on current hardware?
- Others, depending on the particular paper.
- There will be regular meetings (about 6) of small student groups (no more than 3) with the advisor in which you will explain your progress to other meeting participants. The idea behind is to teach you better scientific communication skills. The meetings will be (typically) in German or in English.
- There are two options for presenting your results (will be agreed with your advisor):
- A final talk (about 20 minutes + 5 minutes of discussion) in English language. The slides of the presentation have to be delivered to your advisor. The best talks will be invited to be held in our Konversatorium. This will give the author(s) the chance to present their work to a larger scientific audience and get valuable feedback.
- OR a written scientific report in English language, similar to a research paper. It should exhibit the usual structure like "Abstract, Introduction, .., Conclusions, References", and will be about 8-10 pages.
Downloads
- For the written elaboration: publStyle-Sem186.zip LaTeX EGPubl Class and Author Guidelines. This document class is based on the one used by Eurographics. Two .sty files for our seminars and a Makefile Template are included. The .tex file to be used is called authorGuidelines-sem186-fin-electronic.tex, which includes a similarly named .inc file for the body text.
- PowerPoint-Template for the presentation.
