Intelligent systems through learning algorithms
Machine Learning
Our research centers around two areas in particular: reinforcement learning and representation learning. We also focus on the development of strategies that increase the extent to which data-driven approaches are able to adapt to changing conditions, thus making these approaches a more constant factor than changes. The fields of application include image and language processing, autonomous navigation and recommender systems. We have set out to acquire an in-depth understanding of machine learning, plus its concepts and application scenarios, which is a prerequisite for the efficient deployment of these technologies in real environments.
In order to simplify the transfer of technology, our field of competence operates the “One Stop Shop” Machine Learning Lab (One-ML), where we offer our partners and other interested organizations in industry, training and education and society, a range of services from a single source. These services include technical coaching in the area of machine learning, consulting on various topics - from theory and development to implementation - as well as public presentations outlining the opportunities and risks associated with machine learning.
Open-Source-Tools
Model-based development of safe and cyber-resilient autonomous driving systems
Behaviour modelling and development of autonomous vehicles
Whitepaper
Contact
Projects
Publications
- Laplace Approximation for Real-time Uncertainty Estimation in Object Detection 25th IEEE International Conference on Intelligent Transportation Systems (ITSC), ():, 2022. Details BIB
- SViT: Hybrid Vision Transformer Models with Scattering Transform 32nd IEEE International Workshop on Machine Learning for Signal Processing (MLSP), ():, 2022. Details BIB
- Low-Redundancy Realization of Safety-Critical Automotive Functions In MikroSystemTechnik (MST) Kongress, 2021. VDE Verlag GmbH Berlin Offenbach.. Details URL BIB
- Dynamic Texture Recognition via Nuclear Distances on Kernelized Scattering Histogram Spaces In IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 3755–3759, 2021. Details DOI BIB
- Neural Network and Correlation based Earth-Fault Localization utilizing a Digital Twin of a Medium-Voltage Grid In e-Energy '21: Proceedings of the Twelfth ACM International Conference on Future Energy Systems, pages 249–253, 2021. Details DOI BIB
- Draw with Me: Human-in-the-Loop for Image Restoration In the 43rd German Conference on AI, pages 245-248, 2020. Springer. Details URL BIB
- Towards a Mathematical Understanding of the Difficulty in Learning with Feedforward Neural Networks Proceedings of the 31st IEEE Conference on Computer Vision and Pattern Recognition (CVPR), ():811–820, 2018. Details BIB