392131 From Datasets to Real-World Skills – How Robots Learn, Then Keep Learning (S) (WiSe 2025/2026)

Contents, comment

Robots can learn in two main ways: by imitating demonstrations (imitation learning) or by trial and error (reinforcement learning). Imitation learning enables fast skill acquisition from demonstrations but struggles outside the situations shown in the data. Reinforcement learning can adapt to new situations through interaction, but it is often sample-inefficient and unstable for real robots.

This seminar explores offline-to-online reinforcement learning as a way to combine the strengths of both approaches: starting from large datasets and improving policies through online experience. We will discuss key algorithms for offline RL, such as AWAC, IQL, CQL, recent work showing that standard offline RL algorithms struggle with online fine-tuning, and ways to resolve that. We then look at methods that explicitly integrate imitation learning with reinforcement learning to improve exploration and sample efficiency. Finally, we address continual and lifelong learning, where robots must adapt to changing tasks and environments without forgetting previously acquired skills.

Teaching staff

Dates ( Calendar view )

Frequency Weekday Time Format / Place Period  
weekly Mo 14-16 B2-240 13.10.2025-06.02.2026

Subject assignments

Module Course Requirements  
39-M-Inf-AI-app-foc_a Applied Artificial Intelligence (focus) Applied Artificial Intelligence (focus): Seminar Student information
Applied Artificial Intelligence (focus): anwendungsorientiertes Seminar 1 Study requirement
Student information
Applied Artificial Intelligence (focus): anwendungsorientiertes Seminar 2 Student information
- Ungraded examination Student information
39-M-Inf-ASE-app-foc_a Applied Autonomous Systems Engineering (focus) Applied Autonomous Systems Engineering (focus): Seminar Student information
Applied Autonomous Systems Engineering (focus): anwendungsorientiertes Seminar 1 Student information
Applied Autonomous Systems Engineering (focus): anwendungsorientiertes Seminar 2 Student information
- Ungraded examination Student information
39-M-Inf-INT-app-foc_a Applied Interaction Technology (focus) Applied Interaction Technology (focus) - Seminar Student information
Applied Interaction Technology (focus): anwendungsorientiertes Seminar 1 Study requirement
Student information
Applied Interaction Technology (focus): anwendungsorientiertes Seminar 2 Student information
- Ungraded examination Student information

The binding module descriptions contain further information, including specifications on the "types of assignments" students need to complete. In cases where a module description mentions more than one kind of assignment, the respective member of the teaching staff will decide which task(s) they assign the students.


The seminar will start with two introductory lectures. Each week, one student will present a research paper from the reading list and lead the discussion. At the end of the semester, each student will submit an essay answering a guiding question related to their topic.

No eLearning offering available
Address:
WS2025_392131@ekvv.uni-bielefeld.de
This address can be used by teaching staff, their secretary's offices as well as the individuals in charge of course data maintenance to send emails to the course participants. IMPORTANT: All sent emails must be activated. Wait for the activation email and follow the instructions given there.
If the reference number is used for several courses in the course of the semester, use the following alternative address to reach the participants of exactly this: VST_568321983@ekvv.uni-bielefeld.de
Notes:
Additional notes on the electronic mailing lists
Last update basic details/teaching staff:
Sunday, June 22, 2025 
Last update times:
Wednesday, August 13, 2025 
Last update rooms:
Wednesday, August 13, 2025 
Type(s) / SWS (hours per week per semester)
seminar (S) / 2
Language
This lecture is taught in english
Department
Faculty of Technology
Questions or corrections?
Questions or correction requests for this course?
Planning support
Clashing dates for this course
Links to this course
If you want to set links to this course page, please use one of the following links. Do not use the link shown in your browser!
The following link includes the course ID and is always unique:
https://ekvv.uni-bielefeld.de/kvv_publ/publ/vd?id=568321983
Send page to mobile
Click to open QR code
Scan QR code: Enlarge QR code
ID
568321983