Seminar: Learning for Robot Manipulation

Organizer:
Dr. Tim Welschehold

Co-Organizers:
Imen Mahdi, Nick Heppert

The ability of robotic systems to interact with, grasp, and transform objects in the physical world is one of the most active and challenging frontiers in modern robotics research. Recent advances in machine learning have fundamentally changed how we approach this problem, shifting from explicit programming hand-crafted paradigms toward systems that can acquire complex skills from data, simulation, and own experience. This seminar offers master's students in computer science a structured entry point into this rapidly evolving field, spanning topics from reinforcement and imitation learning to tactile sensing, foundation models, and human–robot interaction. Students will work in groups of three to four, with each group assigned one of twelve carefully selected topics that together cover the breadth of the field. The first task for each group is to independently identify the key sub-topics and open research questions within their assigned area, drawing on recent conference proceedings, journal articles, and survey papers. Students will then conduct a thorough literature review, synthesizing findings across the most relevant and impactful works. Each group will present their topic to the full seminar in an oral presentation, giving their peers an accessible overview of the landscape and the state of the art. In addition, each group will produce a written mini-survey that structures their findings in the style of an academic survey paper, complete with a coherent narrative and proper citations.

Course Information

Details:	Course Number: 11LE13S-7354-M Places: 20 Location: Georges-Köhler-Allee 80, Room Number 00.021
Course Program:	Introduction: 23/04/2026 @ 16:00 How to make a presentation: TBA Block Seminar: 28/07/2026
Evaluation Program:	Seminar Presentation: 28/07/2026   Survey Due Date: 03/08/2026 @ 23:59 CEST
Requirements:	Basic knowledge of Deep Learning or Reinforcement Learning or Robotics
Remarks:	Topics for the seminar will be assigned via preference voting in the form. If there are more interested students than places, places will be assigned based on priority suggestions of the HisInOne system. The date of registration is irrelevant. In particular, we want to avoid that students grab a topic and then leave the seminar. Please have a coarse look at all available topics to make an informed decision before you commit.

Course Material

Google Form:	Topic Selection Form
Slides:	Introduction session How to give a presentation: TBD
Literature Review Guide:	Seminar Literature Review Guide
Templates:	Summary Presentation

Additional Information

Enrollment Procedure

• Enroll through HISinOne, the course number is 11LE13S-7354-M .
• The registration period for the seminars are from 20/04/2026 to 27/04/2026.
• Attend the introductory session on 23/04/2026.
• Fill the Form above for topic selection (will be provide after the place assignment by HISinOne).

Seminar Format Details

• Students will be assigned one topic in groups of 3 or 4.
• Groups will identify relevant subtopics for their assigned topic.
• After confirming with supervisor, subtopics are distributed in group (one per person).
• Students will identify relevant publications for their assigned subtopic.
• Students are expected to prepare a presentation (10 minutes per person) and written survey with their group.
• The seminar will be held as a "Blockseminar" on 28/7/2026
• The slides of your presentation should be discussed with the supervisor two weeks before the Blockseminar.
• The survey should not exceed seven pages (excluding bibliography and images) and is due on 3/08/2026. Significantly longer surveys will not be accepted.
• For details on the seminar structure and evaluation criteria check the slides for the introduction session.

Topics

1. Reinforcement Learning
   Supervisor: Imen Mahdi
2. Imitation Learning
   Supervisor: Nick Heppert
3. Foundation Models in Robot Manipulation
   Supervisor: Imen Mahdi
4. Representation Learning
   Supervisor: Nick Heppert
5. Grasping
   Supervisor: Nick Heppert
6. Dexterous Manipulation
   Supervisor: Imen Mahdi
7. Tactile Sensing in Robot Manipulation
   Supervisor: Tim Welschehold
8. World Models
   Supervisor: Tim Welschehold
9. Sim-2-Real Transfer
   Supervisor: Imen Mahdi
10. Human-Robot Interaction
    Supervisor: Nick Heppert
11. Mobile Manipulation
    Supervisor: Tim Welschehold
12. Multi-modal Manipulation
    Supervisor: Tim Welschehold

Questions?

If you have any questions, please direct them to Imen Mahdi before the topic allotment, and to your supervisor after you have received your topic.