Distributed Intelligence

We continuously offer BSc/MSc theses, research projects and practicals that are closely tied to our current research. Because the field moves fast, we do not publish a fixed topic list. We match motivated students to open problems. Please read the application instructions on the contact page before writing to us.

Application

All applications must follow the instructions on the "How to apply / contact us" page. This page contains the email address and an exact list of what to send.

BSc / MSc Theses

  1. Check your background: math (analysis, linear algebra, probability), programming (Python, PyTorch/JAX), prior ML or computer vision/RL course.
  2. Scan our recent research topics: model-based RL, object-centric self-supervised RL, intrinsic motivation, ML for science, haptic sensing, combinatorial optimisation.
  3. Apply: follow the steps on the “How to apply / contact us” page and indicate “BSc thesis” or “MSc thesis”.
  4. Matching: we assign you to a PhD/PostDoc working on that topic.
  5. Presentations: you present in our monthly student spotlight meeting.

Typical thesis areas (examples, not a fixed list):

  • Data-efficient model-based RL for continuous control
  • Curiosity and intrinsic motivation for embodied agents
  • Object-centric world models for manipulation
  • ML for science: equation learning / symbolic regression
  • Learning-based haptic sensing
  • Neural combinatorial optimisation

Research Projects / Practicals (ML Master)

Intended for students of the ML Master (semester ≥ 2) with good grades.

Research Project ML

  • 9 ECTS, ~270 h
  • conducted individually
  • goal: contribute to a running research line (implement, evaluate, report)

Practical ML

  • 6 ECTS, ~180 h
  • 2–5 students
  • goal: build a software/ML component for an ongoing project

Application: follow the instructions on the "How to apply / contact us" page and use subject “Research Project ML” or “Practical ML”.

Team Projects (BSc)

For BSc Informatik students who want hands-on ML/RL in small groups.

  • 3–5 students
  • supervision by a PhD/PostDoc
  • deliverables: code, short report, final demo to group
  • example projects: curiosity-driven exploration demo, object-centric RL baseline, haptics dataset tooling

To join, follow the instructions on the “How to apply / contact us” page and state “Team Project (BSc)”.