Research Areas
Our key research areas
Automated Machine Learning
Machine learning (ML) is the key technology of our time. We're pushing this to the next level by the progressive automation of ML (AutoML), to democratize access to ML and to realize AI that builds AI.
Neurorobotics
Developing machine learning techniques for robotics and neurotechnology. We focus on data-efficient deep reinforcement learning methods that provide robust and adaptive solutions, and target applications on real-world devices.
Robot Learning
Robots give embodiment to artificial agents thus enable them to affect the physical world. We are researching how these agents can not just automate existing tasks, but be truly autonomous.
Computer Vision
Computer Vision aims to automatically understand visual data with modern approaches heavily relying on deep learning.
Automated Reasoning
Reasoning methods provide guarantees over system behavior which are essential for reliable autonomous agents. We aim to verify learned behavior while also learning to improve our verification methods.
The mission of the ELLIS Unit of the University of Freiburg is to act as one of the best places in Europe on autonomous learning, and to foster European collaborations in the intersection of machine learning, robotics, computer vision and reasoning. The unit has a unique set of outstanding researchers in automated machine learning, robot learning, computer vision, and automated reasoning, and close connections to the industry (e.g., Amazon, Bosch and Toyota). The unit will facilitate collaborations on the intersection of these fields within the ELLIS network, building on highly successful interactions between the respective groups within Freiburg itself. With the recent convergence of much of machine learning, computer vision and robotics under the common theme of deep learning, there are great opportunities for exploiting synergies between these fields.