Credit: Anna Logue
Generative Computer Vision Models
Contents
In this course, we teach fundamental concepts of deep generative models along the example of image generation. Considered models include autoregressive models like RixelRNN and PixelCNN, normalizing Flows and RealNVP-like models, autoencoders (including latent space visualizaiton and deep clustering), variational autoencoders, generative adversarial networks, and diffusion models.
Time and Location
- Lecture: Start: Feb 2026
- Exercise: Start: Feb 2026
Instructor
Participation
- The course is open to students of the Master Business Informatics and Mannheim Master in Data Science (MMDS).
- The course is restricted to 50 participants.
- Places are assigned on first come/
first serve basis.
Outline
Topics include:
- Clustering
- Crash-Course on Neural Networks
- Convoluional Neural Networks,
- RNNs and Transformers
- Autoregressive Models
- Normalizing Flows
- Autoencoders and Latent Space Visualization
- Deep Clustering
- Variational Autoencoders
- Generative Adversarial Networks
- Diffusion Models