Starting this winter term, the chair for Multimedia Computing and Computer Vission is offering the follow-up master lecture Advanced Machine Learning and Computer Vision:

• Master CS or side-subject CS
• 2+2 lecture and tutorial
• Follow-up lecture to Mutlimedia II (INF-0092) / Machine Learning and Compter Vision (INF-0316)

This course is intended for graduate students in CS (or with side-subject CS) and builds on the foundations of our main master course Multimedia II / Machine Learning and Compter Vision. The lecture offers insights into the current development in machine learning, mainly at the border to computer vision. After a introduction to some fundamental techniques in machine learning and optimization (Support Vector Machines, Bipartite Graph Matching, Distance Transforms), it will cover the following topics with related concepts and techniques:

• Current deep learning architectures (ResNeXt, MobileNet, HRNet)
• Embedding Learning, (Visual) Word Embeddings
• Generative Adversarial Networks, Cycle GANs
• Attention mechanisms and Transformer networks
• Semantic and instance segmentation
• Human pose estimation

The tutorial offers the possibility to implement and evaluate the learned techniques.

Please note that the course is a direct follow-up course that requires the successfull completion of our main master course  Mutlimedia II (INF-0092) / Machine Learning and Compter Vision (INF-0316).

For more information, please visit the course page or Digicampus.