Science

Machine Learning, AI, Sound Signal Analyses, Music Instrument Recognition

University studies

I studied computer science and musicology at the Julius-Maximilians-Universität of Würzburg. I was majoring in theoretical computer science. The classification of problems into complexity classes was one of the most interesting topics of my studies.

In my Studienarbeit – the former bachelor’s thesis – I developed an ear training program for special intervals that arise with other tunings. And during my diploma thesis, I used the logical programming language Prolog to analyze Bach chorales (in the MusicXML format) for chords, keys and Co. .

Doctorate

In my doctoral thesis at the international Audio Laboratories Erlangen (a joint institution of the Friedrich-Alexander University Erlangen-Nuremberg and the Fraunhofer IIS), I worked on musical instrument recognition in audio signals conducted research on the optimization of artificial neural networks.

recent research projects

I have been leading the first year of the Artificial Intelligence in Culture and Arts (AICA) group at the University of Music and Performing Arts Munich (HMTM) from April 2023-April 2024, where I am additionally a lecturer for my course application-oriented programming at the chair: Master’s program in Digital Communication.

Doctoral Thesis: Single Musical Instrument Recognition – a solved Problem?

EstherFeeFeichtnerDissertation_fuer_Druck_mit_DanksagungenHerunterladen

Abstract:

Single instrument recognition is a research topic that is worked on for many years. After the AI winter the approaches switched from feature based methods to artificial neural networks. When the recognition rates increased the community moved on to harder problems, at the latest when the community announced that the single instrument recognition is solved.

In this thesis I am going to show that the assumptions that are the basis for this statement are erroneous. The main reason is the misbelief of the comparability of accuracies. To approach this problem I will introduce an evaluation method that is less prone to external factors and a well obtainable data base that enables a better comparison of algorithms.

To evaluate the actual level of the resolving of single instrument recognition and to push this frontier, I refined, combined, and evaluated the feature and the artificial neural network based approach. During this process a new feature was developed and well known features were improved, and artificial neural network architectures were optimized and compared by applying evolutionary programming.

The full thesis: https://open.fau.de/items/f3ff2ed3-3b7a-4222-a69a-8019f23cbce7

Article: A Comparison of Human Creativity and Generative Music Production

Esther_Fee_Feichtner_JahrbuchKulturpolitikHerunterladen

Recent Activities

Discussion panels:

  • 28. June 24 – Expert Panel at „Festival der Zukunft“ in Munic

  • May 15, 2023 – AI_vs: Art – Episode 1 (online)
  • 28.06.2023 – Podiumsdiskussion bei Artistic Intelligence im Leonardo Zentrum Nürnberg
Artistic Intelligence – AI in Culture and Arts

Moderation:

Keynotes:

  • 24. April 24 – „Unleashing Creativity with AI & Sound“ 🎨 at the Data and AI Leaders Forum Munic

Interviews and Potcasts:

  • human Magazin Newsletter:

https://human-magazin.de/?na=v&nk=540-8bbdfee331&id=23

  • XPLR Media:

https://www.xplr-media.com/de/xplr-magazin/ki-in-der-musik-in-manchen-aspekten-wesentlich-kreativer-als-wir.html

  • Radio Interview für WDR COSMO