Journal
SCIENTIFIC REPORTS
Volume 12, Issue 1, Pages -Publisher
NATURE PORTFOLIO
DOI: 10.1038/s41598-022-26429-y
Keywords
-
Categories
Funding
- German Research Council (DFG) [MA-4898/18-1]
- Paul G. Allen Frontier's Group
- Max Planck Society
- Max Planck Society Independent Group Leader Fellowship
- International Max Planck Research School (IMPRS) for Organismal Biology
- German Academic Exchange Service (DAAD)
- United States Department of Defense, Environmental Security Technology Certification Program (ESTCP) [RC 201615]
- Animal Minds Project e.V.
- Carlsberg Foundation Semper Ardens grant
- Danish Environmental Protection Agency
- Dansk Akustisk Selskab
- University of Southern Denmark
- Office of Naval Research
- SDU Lighthouse Project
- Heidelberg Academy of Sciences
Ask authors/readers for more resources
Bioacoustic research often relies on manual identification of target species or call types, which is time-consuming and error-prone. This study presents an open-source deep learning framework, ANIMAL-SPOT, which achieves high accuracy in identifying bioacoustic signals without requiring animal-specific machine learning approaches. The framework is accessible to a broad audience and does not rely on expert knowledge or special computing resources.
Bioacoustic research spans a wide range of biological questions and applications, relying on identification of target species or smaller acoustic units, such as distinct call types. However, manually identifying the signal of interest is time-intensive, error-prone, and becomes unfeasible with large data volumes. Therefore, machine-driven algorithms are increasingly applied to various bioacoustic signal identification challenges. Nevertheless, biologists still have major difficulties trying to transfer existing animal- and/or scenario-related machine learning approaches to their specific animal datasets and scientific questions. This study presents an animal-independent, open-source deep learning framework, along with a detailed user guide. Three signal identification tasks, commonly encountered in bioacoustics research, were investigated: (1) target signal vs. background noise detection, (2) species classification, and (3) call type categorization. ANIMAL-SPOT successfully segmented human-annotated target signals in data volumes representing 10 distinct animal species and 1 additional genus, resulting in a mean test accuracy of 97.9%, together with an average area under the ROC curve (AUC) of 95.9%, when predicting on unseen recordings. Moreover, an average segmentation accuracy and F1-score of 95.4% was achieved on the publicly available BirdVox-Full-Night data corpus. In addition, multi-class species and call type classification resulted in 96.6% and 92.7% accuracy on unseen test data, as well as 95.2% and 88.4% regarding previous animal-specific machine-based detection excerpts. Furthermore, an Unweighted Average Recall (UAR) of 89.3% outperformed the multi-species classification baseline system of the ComParE 2021 Primate Sub-Challenge. Besides animal independence, ANIMAL-SPOT does not rely on expert knowledge or special computing resources, thereby making deep-learning-based bioacoustic signal identification accessible to a broad audience.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available