We permanently offer proposals for bachelor and master theses topics; take a look at the list of topics below or at our staff's pages to see what they do and contact whoever you'd like to work with directly. Please note that:
- You need to be a student of RWTH Aachen University. Again: we do NOT offer theses to non-RWTH students.
- You need prior experience in computer vision, such as e.g. having taken part in any of our classes.
- You should have solid programming experience. Prior experience with OpenCV is a plus.
- We do NOT offer any internships. Please do NOT contact us for an internship.
Tracking multiple objects in a scene, given potential detections boils down to a data association problem, where a promising method is Markov Chain Monte Carlo Data Association (MCMCDA). Still, the pure data association does not solve the problems. Extensions are needed to cope with occlusions, mutual exclusions, highly dynamic behavior or social groups of people. Your job would include the work with the (existing) MCMCDA framework and the exploration of those extensions.
Pre-Hiwi possible (and preferred).
For more info contact me! :)
M.Sc. Stefan Breuers
Scene Understanding, SLAM, Localization and Mapping, 3D Reconstruction
incl. Deep Learning Approaches
Dr. Jörg Stückler
Landmark Discovery and Recognition
M.Sc. Theodora Kontogianni