You can find details of current topics for our Digital Health and Computational Intelligence seminars on Digicampus. Please contact one of our  researchers if you have any questions.

We currently have the following topics available for supervision for Bachelor’s and Master’s theses and modules. If you have questions or would like to enquire about other project topics, please get in touch. You can find our research interests on our individual pages.

• Vorhersage von Ausfällen bei Starterbatterien (Masterarbeit mit MAN Truck & Bus SE) (, )
• Emotional text-to-speech generation ( Shahin Amiriparian)
• Meta learning for multi-corpus speech analysis ( Maurice Gerczuk)
• Federated learning for mobile-based audio analysis ( Maurice Gerczuk)

• Encoded speech actvity recognition in noisy environments ( Manuel Milling, Andreas Triantafyllopoulos)

• Detection and direction of arrival estimation for sound events ( Manuel Milling, Andreas Triantafyllopoulos)

• Exploring optimisation algorithms for deep learning ( Manuel Milling)
• Object detection and classification of pollen on microscope images ( Manuel Milling)
• Exploring distance-based loss functions for deep learning in classification tasks ( Manuel Milling)
• Multi-modal clinical score prediction for multiple sclerosis ( Manuel Milling, Alexander Kathan)
• Exploring cross-modal interactions for emotion recognition in the wild ( Vincent Karas)
• Exploring the effectiveness of transfer learning for emotion detection ( Vincent Karas)
• Audio-visual acoustic scene classification ( Andreas Triantafyllopoulos)
• Runner wellbeing and fatigue prediction from wearable and audio data ( Andreas Triantafyllopoulos)
• Bio-acoustic event detection ( Andreas Triantafyllopoulos)
• Personalised humour recognition using linguistic features ( Alexander Kathan)
• A personalised multimodal approach to humour recognition ( Alexander Kathan)
• Exploring personalised machine learning in healthcare applications ( Alexander Kathan)
• A novel similarity-based personalisation approach for speaker-independent audio processing ( Alexander Kathan)
• Clinical score prediction for multiple sclerosis  utilising personalised models  ( Alexander Kathan, Manuel Milling)