Thomas
Dethmann
ML Research Engineer | Musician
Audio enthusiast at the intersection of music, acoustics, and deep learning.
01 / About
Background
I hold two Bachelor's degrees — one in English and American Studies & Musicology, and another in Audio and Video, which included a semester abroad at LTU in Southfield, MI studying Audio Engineering.
I then pursued a Master's degree in Audio Communication and Technology, during which I worked as a Machine Learning Engineer Intern at Apple. My Master's thesis, Automatic Drum Transcription with Deep Neural Networks, was supervised by Prof. Weinzierl and Andreas Arzt.
In my spare time, I'm the drummer of Fil der Protagonist.
Machine Learning
Deep neural networks for audio signal processing and automatic transcription.
Audio & Acoustics
Signal processing, sound synthesis, and audio communication technology.
Research
Master's thesis on Automatic Drum Transcription with Deep Neural Networks.
Musician
Drummer of Fil der Protagonist. Three years of formal drum studies.
02 / Work
Projects
Automatic Drum Transcription
Deep learning-based automatic drum transcription (ADT) system. At first, you hear the originally recorded drums followed by the ADT version, where all subtypes of drum hits are mapped to Kick, Snare, Tom, Hi-Hat, Crash, Ride, and Bell. Finally, the original transitions to the ADT version for direct comparison.
COVID-19 Sonification
In order to better understand the changes of COVID-19, we implemented a sonification of the RKI COVID-19 dataset using Python and Faust.
RTAP Synthesizer
A functioning Pure Data synthesizer patch with an additional external written in C. Developed as a collaborative student project.
03 / Music
Solo / Band
Final Exam
During my Bachelor's in Audio and Video, I studied drums for 3 years. This improvised solo was performed as part of my final exam.
Fil der Protagonist
Band performances and releases with Fil der Protagonist.
04 / Contact
Get in Touch
Interested in collaborating on ML, audio, or music tech projects? Feel free to reach out.