Birds, bats and beyond: Evaluating generalization in bioacoustics models

B Van Merriënboer, J Hamer, V Dumoulin… - Frontiers in Bird …, 2024 - frontiersin.org
In the context of passive acoustic monitoring (PAM) better models are needed to reliably
gain insights from large amounts of raw, unlabeled data. Bioacoustics foundation models …

Global birdsong embeddings enable superior transfer learning for bioacoustic classification

B Ghani, T Denton, S Kahl, H Klinck - Scientific Reports, 2023 - nature.com
Automated bioacoustic analysis aids understanding and protection of both marine and
terrestrial animals and their habitats across extensive spatiotemporal scales, and typically …

Out-of-domain robustness via targeted augmentations

I Gao, S Sagawa, PW Koh… - International …, 2023 - proceedings.mlr.press
Abstract Models trained on one set of domains often suffer performance drops on unseen
domains, eg, when wildlife monitoring models are deployed in new camera locations. In this …

Domain‐specific neural networks improve automated bird sound recognition already with small amount of local data

P Lauha, P Somervuo, P Lehikoinen… - Methods in Ecology …, 2022 - Wiley Online Library
An automatic bird sound recognition system is a useful tool for collecting data of different
bird species for ecological analysis. Together with autonomous recording units (ARUs), such …

Extensive data engineering to the rescue: building a multi-species katydid detector from unbalanced, atypical training datasets

S Madhusudhana, H Klinck… - … Transactions of the …, 2024 - royalsocietypublishing.org
Passive acoustic monitoring (PAM) is a powerful tool for studying ecosystems. However, its
effective application in tropical environments, particularly for insects, poses distinct …

Automatic detection for bioacoustic research: a practical guide from and for biologists and computer scientists

A Kershenbaum, Ç Akçay, L Babu‐Saheer… - Biological …, 2024 - Wiley Online Library
Recent years have seen a dramatic rise in the use of passive acoustic monitoring (PAM) for
biological and ecological applications, and a corresponding increase in the volume of data …

Feature embeddings from large-scale acoustic bird classifiers enable few-shot transfer learning

B Ghani, T Denton, S Kahl, H Klinck - arxiv preprint arxiv:2307.06292, 2023 - arxiv.org
Automated bioacoustic analysis aids understanding and protection of both marine and
terrestrial animals and their habitats across extensive spatiotemporal scales, and typically …

A dataset for benchmarking Neotropical anuran calls identification in passive acoustic monitoring

JS Cañas, MP Toro-Gómez, LSM Sugai… - Scientific Data, 2023 - nature.com
Global change is predicted to induce shifts in anuran acoustic behavior, which can be
studied through passive acoustic monitoring (PAM). Understanding changes in calling …

Unsupervised classification to improve the quality of a bird song recording dataset

F Michaud, J Sueur, M Le Cesne, S Haupert - Ecological Informatics, 2023 - Elsevier
Open audio databases such as Xeno-Canto are widely used to build datasets to explore bird
song repertoire or to train models for automatic bird sound classification by deep learning …

soundscape_IR: A source separation toolbox for exploring acoustic diversity in soundscapes

YJ Sun, SC Yen, TH Lin - Methods in Ecology and Evolution, 2022 - Wiley Online Library
Soundscapes contain rich acoustic information associated with animal behaviours,
environmental characteristics and human activities, providing opportunities for predicting …