Curriculum Vitae

📄 One-page résumé as PDF

I’m looking for exciting new challenges at the intersection between machine learning, voice technology and art, between scientific theory and applications in creative domains.

Research Interests

AI for voice, music, and sound:

Open to new challenges related to artistic media.

Experience

Research Scientist
Supertone Inc. • Korea
2024 - today
  • Develop, train, and package Supertone’s voice backbone NANSY
    • Improve controllability through improved pitch encoder
    • Optimize model architecture for better synthesis quality and faster inference
  • Experiment with LLM-based TTS and neural audio codecs
  • Develop and maintain voice editing tool built on Supertone’s voice technology
  • Publish and present research at international conferences
Post-doc, Doctoral researcher, Intern
IRCAM • France
2018 - 2024
  • Research on deep learning for voice
    • Topics: Neural vocoding and representation, self-supervised learning, voice attribute transformation
  • Technical support for film, music, and art: bespoke voice transformation for artists
  • Application development: research results into voice editing and live transformation apps
Intern
Siemens AG • Germany
2015
Teaching assistant
TU Darmstadt, Signal Processing Group • Germany
2014

Education

Doctor of Philosophy (PhD)
Sorbonne Université • France
2019 - 2023
Master of Science (MSc) - Computational Engineering
TU Darmstadt • Germany
2015 - 2019
Bachelor of Science (BSc) - Electrical Engineering
TU Darmstadt • Germany
2012 - 2014

Additional Education

Additional credits - Mathematics & Music
TU Darmstadt • Germany
2013 - 2018
Study exchange
Tongji University • Shanghai, China
2014 - 2015
Study exchange
AC Davis High School • Yakima, WA, USA
2009 - 2010

Skills

Research: Experimental design and evaluation · scientific writing; presented at ICASSP, Interspeech, ISMIR

Machine learning: Generative modeling · self-supervised learning · neural codecs and vocoders

Domain knowledge: Signal processing · acoustics · voice science · music theory

Engineering: Python (PyTorch, NumPy) · application development · model deployment

Languages: German (native) · English (C2) · French (B2) · Polish (B1) · Chinese (A2) · Korean (basic)

Music: Classically trained singer · composition · live coding (Tidal Cycles) · piano

Scholarships

Full PhD scholarship • EDITE de Paris (ED130)
2019 - 2022
Full study scholarship • Studienstifung des deutschen Volkes (SdV)
2015 - 2019
Deutschlandstipendium • TU Darmstadt, Inro Elektrotechnik, Siemens
2013 - 2014

Activities

Student representative for HCERES evaluation
2023
Founded a live coding ensemble with scientific staff at IRCAM
2021 - 2023
SdV Summer Academy on Impact on video streaming on film industry
2018
Orchestral composition Das Mädchen aus Hunan
2017 - 2018
Improvisation ensemble at Akademie für Tonkunst Darmstadt, Piano and Gyrophon
2016 - 2018
Comedian harmonists singing quartet at Staatstheater Darmstadt
2016 - 2018
Choir member Chor der TU Darmstadt
2012 - 2018
SdV Summer Academy on soundtrack and soundscape
2017
Composition at SdV Music Academy
2016 - 2017
Choir director Mathechor
2013 - 2014
Student representative at electrical engineering department of TU Darmstadt
2012 - 2014

Talks

Circe, the IRcam neural auto-encoder for voiCE • IRCAM Forum at NYU
2022
Quell'usignolo che innamorato, the resurgence of an artificial voice • Mercredis de STMS
2022
Workshop on neural vocoding • Deep Voice Paris
2021

Code

Checkout my github page for code examples, though, unfortunately, most of my actual code cannot be made available.