BSc / MSc Theses

How do I write my BSc / Msc Thesis in the Autonomous Vision Group?

We continuously offer BSc and MSc theses in the context of our current research. As our field advances quickly, we do not maintain a list of thesis projects but determine these topics on demand. If you are interested in doing your BSc or MSc thesis with us, please follow these steps:

  • You need to have a solid math (analysis, algebra, statistics) and programming (python, PyTorch) background and prior experience in machine learning, computer vision or both. If you don't, please take the relevant courses before approaching us.
  • Familiarize yourself with our latest research (last 2 years). You find all abstracts of our recent publications on this page. Read the papers that you find most interesting in detail and write down your comments or think about potential extensions.
  • Send an email to Prof. Andreas Geiger that describes your programming experience, your research interests and your project ideas, if you have any. Mention the papers from us that you find most interesting and send us your comments/ideas. Please also attach your CV and transcripts.

We will then match your interest to one of our research projects.

Project Proposals

While we typically create research projects based on students interest, sometimes we also have concrete project proposals listed here.


Past BSc/MSc Theses

  • Markus Flicke (MSc): 
  • Yifan Zhu (MSc): Learning 3D Shape Completion under Self-Supervision
  • Siyuan Peng (BSc): Semantic Scene Understanding on KITTI-360
  • Simon Doll (MSc): Lost cargo detection using sensor fusion and machine learning
  • Nico Widmann (BSc): The Use of Image Search Engines for Mining Implicitly Labelled Datasets for Image Classification
  • Maximilian Beller (MSc): Novelty Detection in High Dimensional and Sparse Spaces
  • Danilo Brajovic (MSc): LIDAR Appearance Features for 3D Object Tracking in Autonomous Driving
  • Paul Sanzenbacher (MSc): Learning a Photorealistic Renderer
  • Stefan Bergmann (MSc): Learning Material Reflectance for Photorealistic Rendering
  • Arne Gebert (BSc): Active Depth Reconstruction 
  • Johannes Zenn (BSc): Unsupervised Image-to-Image Translation with Induced Feature Statistics
  • Marius Hobbhahn (BSc): Object Detection using the Scattering Transform
  • Paul Wullenweber (BSc): Calibrating A New 3D Scanner
  • Kuan-Chun Lee (MSc): On Representation Learning with Autoencoding Neural Networks
  • Marissa Weiss (MSc): Class-agnostic Instance Segmentation with Foveated Image Sampling
  • Robert Herzig (MSc): Supervised Single Image Depth Estimation with Deep Convolutional Neural Network