Thorsten Eisenhofer is a tenure-track faculty at CISPA Helmholtz Center for Information Security in Saarbrücken, Germany. He previously worked as a postdoctoral researcher at BIFOLD and TU Berlin and earned his PhD from Ruhr University Bochum, where he was part of the Cluster of Excellence CASA. His research focuses on machine learning and computer security, particularly on attacks against learning-based models and defenses to improve their robustness. This work often involves looking beyond the model itself and examining the entire computational pipeline, including the underlying hardware and software stack.
International Conference on Machine Learning (ICML) Leveraging Frequency Analysis for Deep Fake Image Recognition