International Bioacousics Congress 2025
Vincent Sirius Kather
Researching animal sound using deep learning
How can we discover new animal sounds?
Can we use systems trained on known species to help us find unknown ones?
My research is aimed at investigating, if and how bioacoustic deep learning can
help us find new sounds and perhaps even new species. This is a very relevant question from both
ecology and computer science.
From an ecological perspective, automated monitoring of biodiversity is currently
limited to animal species we have a lot of training data from - and to environments with low
amount of background noise. However, the
biodiversity crisis
we are currently facing is especially
critical for rare and unknown species in habitat's where our automated monitoring strategies currently still
yield low performance.
From a computer science perspective, deep learning models still
act as black boxes, and we don't really know what they listen for.
Luckily, many active researchers in this field are working on these topics. I therefore work on evaluating existing bioacoustic deep learning models and finding out what exactly they are focusing on when organizing large amounts of unlabeled recordings. The goal is to develop an analysis workflow that allows us to find new biological sounds in large amounts of audio recordings using the models we already have.
I am currently enrolled in a PhD project as part of the Marie Skłodowska Curie doctoral network Bioacoustic AI with the project title Identifying unknown species and unknown sounds using machine learning. I am based at the Naturalis Biodiversity Center and enrolled as a PhD student at Tilburg University.
Research Interests
- Deep learning: Feature extraction and classification using deep learning models.
- Evaluation of high-dimensional spaces: Evaluating embedding spaces using clustering and probing.
- Ecoacoustics: Assessing how biodiversity can be quantified using automatic processing of audio signals.
- Connecting bioacoustics and computer science: Making state-of-the-art deep learning models accessible for ecologists and conservationists through interactive tools.
Publications
- bacpipe: a Python package to make bioacoustic deep learning models accessible. Vincent S. Kather, Sylvain Haupert, Burooj Ghani, Dan Stowell. (2026). ArXiv.org. [DOI]
- Optimizing Machine Learning for the Detection of Splashing Sounds for Automatic Monitoring of Fish Spawning. Line Weiss, Lisa Dörner, Vincent S. Kather, Max de Koning, Kees te Velde, Ine Pauwels, Christian Tudorache, Hans Slabbekoorn. (2026). Preprint/Other. [DOI]
- Clustering and Novel Class Recognition: Evaluating Bioacoustic Deep Learning Feature Extractors. Vincent Kather, Burooj Ghani, Dan Stowell. (2025). Preprint/Other. [DOI]
- Uncertainty Calibration of Multi-Label Bird Sound Classifiers. Raphael Schwinger, Ben McEwen, Vincent S. Kather, René Heinrich, Lukas Rauch, Sven Tomforde. (2025). ArXiv.org. [DOI]
- Development of a machine learning detector for North Atlantic humpback whale song. Vincent Kather, Fabian Seipel, Benoît Bergès, G. E. Davis, Catherine A. Gibson, M J Harvey, Lea‐Anne Henry, Andrew Stevenson, Denise Risch. (2024). The Journal of the Acoustical Society of America. [DOI]
- Interaction of equivalence ratio fluctuations and flow fluctuations in acoustically forced swirl flames. Vincent Kather, Finn Lückoff, Christian Oliver Paschereit, Kilian Oberleithner. (2021). International Journal of Spray and Combustion Dynamics. [DOI]
Software
- acodet ★ 12
Repository: Python package for inference and retraining of bioacosutic deep learning models
Posters
- BioacousticAI project Meeting April 2026
- European Cetacean Society Conference 2023
Other Resources
- Presentation at IRCAM 13/05/2026
- Presentation from GloBat 14/04/2026