Practical Course Computer Graphics and Visualization (SS2019)

Supervisor

CrowdGame for Machine Learning

The aim of the practical course is to develop a framework for crowd sourcing in which the ground truth information is obtained in a playful way. In order to capture this information, a Virtual Reality (VR) game will be developed. The approach is intended mainly for learning complex models (especially neural networks) on large amounts of data with minimal manual effort.

Applications:

Semantic segmentation of topographic maps
Cell segmentation (biology)
and many more

Modules

In addition to the modules listed in your study regulations, the practical course can also be included in INF-VERT7 upon informal application to Prof. Gumhold.

Requirements

Understanding of Computer Graphics & Game Engines (Unreal) advantageous
Knowledge of Python (especially PyTorch for neural networks) advantageous

Organization

Enrolment via OPAL. The first meeting will take place in the week from 08 to 12 April 2019. The specific date will be agreed with all participants.