Self Supervised Learning

This is an advanced MSc course on Self Supervised Learning for Summer 2026, designed for 2nd and 3rd semester students.

eCampus

BASIS

Prerequisites

This course requires a good understanding of deep learning and machine learning.

Schedule

Tutorials: Mondays, 16:00–18:00, HS1

Lectures: Fridays, 14:00–16:00, HS4

Date Time Topic Material
Lecture 1 April 17, 2026 14:15–16:00, HS4 Intro slides
Lecture 2 April 24, 2026 14:15–16:00, HS4 Clustering slides sklearn user guide
Bishop Chapter 9.1
kmeans++
Deep Clustering
STEGO
CRFs
Tutorial 2 April 27, 2026 16:15–18:00
May 1, 2026 Public Holiday
Lecture 3 May 8, 2026 14:15–16:00, HS2 Autoencoders Deep Learning Book
Tutorial 3 May 11, 2026 16:15–18:00
- May 15, 2026 - No Lecture
Lecture 4 May 22, 2026 14:15–16:00, HS2 Generative Models
May 29, 2026 ascension break
Tutorial 4 June 1, 2026 16:15–18:00
- June 5, 2026 - No Lecture
Lecture 5 June 12, 2026 14:15–16:00, likely online Denoising
Tutorial 5 June 15, 2026 16:15–18:00
Lecture 6 June 19, 2026 14:15–16:00, HS2 Pretext Augmentations
Tutorial 6 June 22, 2026 16:15–18:00
Lecture 7 June 26, 2026 14:15–16:00, HS2 Student-teacher Learning
Tutorial 7 June 29, 2026 16:15–18:00
Lecture 8 July 3, 2026 14:15–16:00, HS2 Cycle Consistency & Contrastive Learning
Tutorial 8 July 6, 2026 16:15–18:00
Lecture 9 July 10, 2026 14:15–16:00, HS2 Pseudo Labels
Tutorial 9 July 13, 2026 16:15–18:00
Lecture 10 July 17, 2026 14:15–16:00, HS2 System-Level Self-Supervision
Tutorial 10 July 20, 2026 16:15–18:00
Lecture 11 July 24, 2026 14:15–16:00, HS2 Research Insights

FAQ

Mobile Robotics and other Study Programs

Students of the Mobile Robotics MSc program or other programs that are not from the CS department can enroll into this course. Please contact your respective exam office and only contact the lecture team with clear information what your exam office requires. Please also take note of the course prerequisits listed above.

Course Registation

You need to register through BASIS by end of April. Any person registered through BASIS should automatically also be added on eCampus. There is no other way to register for the lecture. Registration for the lecture is a prerequist to later register for the exam.

Problems with BASIS or eCampus

Anybody who is registered on BASIS should be automatically be added to eCampus. If that has not worked or you have technical problems registering on BASIS, please contact bama+tech@informatik.uni-bonn.de.

For any other questions regarding this course, please ask in the forum on eCampus. Only in urgent matters, contact blumh@uni-bonn.de.