8โ€“12 Sept 2026
Europe/Vienna timezone

Session

A02.01 Machine Learning in Bioacoustics

A02.01
8 Sept 2026, 13:40

Conveners

A02.01 Machine Learning in Bioacoustics: S008

  • Jure Zeleznik (Acoustics Research Institute, Austrian Academy of Sciences)
  • Daniel Haider (Acoustics Research Institute, Austrian Academy of Sciences)

A02.01 Machine Learning in Bioacoustics: P423

  • Daniel Haider (Acoustics Research Institute, Austrian Academy of Sciences)
  • Jure Zeleznik (Acoustics Research Institute, Austrian Academy of Sciences)

A02.01 Machine Learning in Bioacoustics: S252

  • Daniel Haider (Acoustics Research Institute, Austrian Academy of Sciences)
  • Jure Zeleznik (Acoustics Research Institute, Austrian Academy of Sciences)

Presentation materials

There are no materials yet.

  1. Daniel Haider (Acoustics Research Institute, Austrian Academy of Sciences)
    08/09/2026, 13:40
    A02 Bio-acoustics

    We present a pipeline for extracting and analyzing the fundamental frequency (F0) contour patterns in the rumbles of African savanna elephants (Loxodonta africana). Rumble F0s can go as low as 15 Hz and are often masked by infrasonic background noise or overlap with other calls, making standard contour extraction methods perform poorly. To address this, we introduce a custom infrasonic-adapted...

    Go to contribution page
  2. Sercan Alipek (University of Siegen)
    08/09/2026, 14:00
    A02 Bio-acoustics

    Automatic detection and classification of bat echolocation calls is essential for large-scale biodiversity monitoring, yet remains challenging under real-world acoustic conditions and constrained hardware environments. This work presents a lightweight and deployable pipeline for bat call analysis that combines simple signal processing with efficient time-series machine learning. An adaptive...

    Go to contribution page
  3. Florian Krebs (Joanneum Research Forschungsgesellschaft mbH)
    08/09/2026, 14:00
    A02 Bio-acoustics

    Large-scale bioacoustic datasets are typically weakly labelled, assigning species labels to entire recordings without temporal annotations. When recordings are divided into short audio chunks for training, many chunks inherit the recording-level label despite containing no target species, introducing weak-label noise. In this work, we quantify this source of noise in the InsectSet459 dataset...

    Go to contribution page
  4. Ignasi Nou-Plana (La Salle - URL)
    08/09/2026, 14:00
    A02 Bio-acoustics

    Effective management of respiratory diseases in cattle (Bos taurus), particularly within the context of Bovine Respiratory Disease Complex (BRDC), relies on early, non-invasive monitoring strategies to safeguard animal health and welfare. Among the earliest observable indicators, vocal and respiratory-related sounds such as coughs and vocalizations provide valuable insight into physiological...

    Go to contribution page
  5. Clara Hollomey (Univ. of Appl. Sciences St. Pรถlte)
    08/09/2026, 14:00
    A02 Bio-acoustics

    We investigate whether the frame-theoretic properties of filterbanks can predict downstream neural network classification performance. Using a dataset of 3,080 African savanna elephant rumble vocalisations classified into four age groups, we train a convolutional neural network on magnitude spectrograms derived from 15 filterbank configurations spanning equivalent rectangular bandwidth (ERB),...

    Go to contribution page
  6. Stephen Marsland (Victoria University of Wellington)
    08/09/2026, 14:20
    A02 Bio-acoustics

    Automated bird call recognition systems trained on one recording environment often produce Automated bird call recognition systems trained on one recording environment often produce worse results when deployed to others due to differences in background noise and recording equipment, a problem known as domain shift. To isolate this issue we process data from two different recording sources into...

    Go to contribution page
  7. Zsofia Katona (Acoustics Research Institute, OEAW)
    08/09/2026, 14:40
    A02 Bio-acoustics

    Artificial intelligence holds powerful potential for studying animal vocal communication in naturalistic group environments, where manual annotation of audiovisual data is often prohibitively time-consuming. We are developing a pipeline to separate individual calls within a group of budgerigars (Melopsittacus undulatus) to investigate potential language-like structure in their vocalizations....

    Go to contribution page
  8. Guillaume Dutilleux (NTNU/IE/IES/Acoustics)
    08/09/2026, 15:20
    A02 Bio-acoustics

    In the context of the decline of the European common spadefoot toad (Pelobates fuscus), long term bioacoustic monitoring of this secretive species that vocalizes underwater is highly relevant. In this study, we present a software detector that features superior performance compared to literature [1]. For this study, hydrophone data was collected in Denmark and Poland with a different recorder...

    Go to contribution page
  9. Guillaume Dutilleux (NTNU/IE/IES/Acoustics)
    08/09/2026, 15:40
    A02 Bio-acoustics

    In Scandinavia, the Eurasian lynx (\textit{Lynx lynx}) is a hunted species whose populations must be monitored. We present a deep-learning-based software detector of E. lynx calls. It can be applied to long-term audio recordings carried out during the breeding season where the species produce loud and far-reaching calls. To our knowledge, this is the first published software detector for this...

    Go to contribution page
  10. Vincent S. Kather (Naturalis Biodiversity Center)
    08/09/2026, 16:00
    A02 Bio-acoustics

    Natural soundscapes, especially in hyper-diverse environments often contain unknown sound events. Strategies to identify unknown sound events, especially those produced by species not included in training sets have traditionally been infeasible to manually identify. Recently, the classification performance of bioacoustic deep learning models is becoming competitive with human annotators. Yet,...

    Go to contribution page
  11. Bram Cuyx (Flanders Marine Institute (VLIZ))
    08/09/2026, 16:20
    A02 Bio-acoustics

    Marine soundscapes contain valuable information about local ecosystems, making comprehensive analysis of their constituent events essential for accurately assessing ecosystem health. However, they are subject to high levels of continuous noise, hindering our ability to disentangle these events and analyse the soundscape. One potential approach to mitigating noise is the use of a...

    Go to contribution page
Building timetable...