Voice Biomarkers for Stroke Diagnosis and Therapy

  • Description: Stroke rehabilitation is a complex process consisting of numerous treatment modules. In this work, we aim at diagnosing possible stroke-related language pathologies (e.g. aphasia, dysarthria, apraxia) by analyzing spoken language. Resulting symptoms should be detected with high accuracy. Based on this, appropriate mobile treatment procedures are to be developed and tested on patients.
  • Term: since July 1, 2021
  • Funding source: Siemens Healthineers (d.hip)

Smart Operation Room Speech Assistant

  • Description: While previous researches in intelligent smart operation room assistance systems investigate signal and video clip-based surgical activity detection, a new kind of voice-based approach is able to unite all available information.
  • Term: since May 1, 2021
  • Funding source: Siemens Healthineers (d.hip)

Bachelor & Master supervisions:


– Christopher Popp (Ms.C): Domain-Specific ASR for German Automotive Domain, 2023.
– Matteo Torcoli (Dr.-Ing.): Dialogue Enhancement and Personalisation – Contributions to Quality Assessment and Control, 2023.
– Nina Brolich (Bs.C): Automatic Pathological Speech Intelligibility Assessment, 2023.
– Sai Varun Varanasi (Ms.C): STApp: Stroke Therapy Application Development using Speech Disentanglement Analysis, 2023.


– Nadine Rucker (Dr.-Ing.): Log File Processing – Changes, Challenges, and Chances, 2022.
– Celine Pohl (Bs.C): Self-supervised learning for pathology classification, 2022.
– Maximilian Riehl (Bs.C): Development and Evaluation of Transformer-based Deep Learning Model for 12-lead ECG Classification, 2022.
– Luis Schmid (Bs.C): Localisation of Ischemic Heart Disease using Machine Learning Methods with Magnetocardiography Data, 2022.


– Sungjae Cho (Ms.C): Simulating Problem Difficulty in Arithmetic Cognition Through Dynamic Connectionist Models, 2019, whose work resulted in international conference publication, and started as a researcher in KAIST (Korea Advanced Institute of Science and Technology).


– Yu Seoha (Ms.C): Longitudinal study of cochlear implantation infants’ development of articulation, 2018, and started a PhD in SNU, 2018.
– Jooyoung Lee (Ph.D): Age Classification from Speech using LSTM, whose work resulted in international conference publication, and started a PhD in SNU, 2018.
– Jongin Kim (Ph.D): Automatic classification of classical music based on musical contents, and started a PhD in SNU, 2018.
– Sabaleuski Matsvei (Ms.C): Vowel Duration and Fundamental Frequency Prediction for Automatic Transplantation of Native English Prosody onto Korean-accented Speech, and started a career as a computational linguist in Megputer in Russia, 2018.
– Evgenia Nedelko (Ms.C): Automatic phrase break and sentence stress prediction in English with RNN, 2018.