Interdisciplinary Project

Master students in computer science can replace the external module by an interdisciplinary project with 18 ECTS credits. The Computer Vision Group and the Image Analysis Group support interdisciplinary projects with the following partners (the list may grow in the future):

These partners may (but are not obliged to) offer projects on problems in their field that require a strong computer science background. Currently the following projects are offered:

  • Analysis of High Speed 3D Imaging Data
    We are interested in the neural circuits underlying locomotor behavior in larval zebrafish. For this purpose, we use high-speed imaging (1000 fps) to resolve the undulatory tail movements. We developed an algorithm in Matlab in which tail curvature is represented as a number of points (e.g. 30) and we would be interested in developing a fast algorithm that classifies behavioral modules (each lasting about 200 ms) according to e.g. the point of maximal tail curvature and the direction of tail beating (turns). The algorithm should preferentially be developed in Matlab, although other programming environments (e.g. C) might be used to increase speed.
The interdisciplinary project is usually more challenging than the external module in the sense that it requires the ability to apply computer science knowledge in a different field without much help from a senior person in computer science. Students must be well organized, communicative, open minded, and capable to work independently. The interdisciplinary project is recommended only to students who bring these abilities. At least two computer science courses relevant for the project must have been taken (e.g. Image Processing, Computer Vision I, 3D Image Analysis). The partner in the other department, Thomas Brox or Olaf Ronneberger, and the examination board must give their agreement before the project can start.

Contact Thomas Brox or Olaf Ronneberger, if you are interested in such a project.