Workshop with CARIAD
The afternoon began with an introduction round which was followed by three overview talks:
-
Tsung Yuan Tseng introduced us to his previous work at the TU Munich, giving insight into his research background and laying the groundwork for his future contributions to the lab.
-
Next, Dr. Matthias Rottmann provided an overview of current research topics at the University of Wuppertal, highlighting projects that align with safe and robust machine learning, especially in the context of computer vision and autonomous systems.
-
Following that, Dr. Fabian Hüger gave an engaging talk on Safe AI Engineering, emphasizing the importance of trustworthy and transparent AI in automotive applications and the challenges that arise in real-world deployment scenarios.
After a lively coffee break and open discussions, the workshop continued with two research presentations from lab members:
-
Edgar Heinert presented joint work on "Transferring Styles for Reduced Texture Bias and Improved Robustness in Semantic Segmentation Networks", developed in collaboration with Ben Hamscher, Annika Mütze, Kira Maag and Matthias Rottmann. The talk focused on improving generalization in vision models through style transfer techniques.
-
Dr. Annika Mütze followed with a talk titled "Can We Challenge Open-Vocabulary Object Detectors with Generated Content in Street Scenes?", co-authored with Sadia Ilyas, Christian Dörpelkus, and Matthias Rottmann. She explored how synthetic data can help test the limits of object detectors trained on broad, open-vocabulary datasets.
This workshop marks a promising step forward in fostering long-term collaboration between academia and industry. We look forward to supporting Tsung Yuan Tseng in his research and continuing to explore the intersection of safe AI and automotive innovation.
