CoChat
Sign Up
Cited by 905Music and Audio ProcessingAnimal Vocal Communication and BehaviorSpeech Recognition and Synthesis

BirdNET: A deep learning solution for avian diversity monitoring

Open access
Kahl, Stefan, Wood, Connor M., Eibl, Maximilian, Klinck, Holger·2021·Ecological Informatics
DOI 10.1016/j.ecoinf.2021.101236
Open this paper in CoChat — free accountOpen PDF

Abstract

Variation in avian diversity in space and time is commonly used as a metric to assess environmental changes. Conventionally, such data were collected by expert observers, but passively collected acoustic data is rapidly emerging as an alternative survey technique. However, efficiently extracting accurate species richness data from large audio datasets has proven challenging. Recent advances in deep artificial neural networks (DNNs) have transformed the field of machine learning, frequently outperforming traditional signal processing techniques in the domain of acoustic event detection and classification. We developed a DNN, called BirdNET, capable of identifying 984 North American and European bird species by sound. Our task-specific model architecture was derived from the family of residual networks (ResNets), consisted of 157 layers with more than 27 million parameters, and was trained using extensive data pre-processing, augmentation, and mixup. We tested the model against three independent datasets: (a) 22,960 single-species recordings; (b) 286 h of fully annotated soundscape data collected by an array of autonomous recording units in a design analogous to what researchers might use to measure avian diversity in a field setting; and (c) 33,670 h of soundscape data from a single high-quality omnidirectional microphone deployed near four eBird hotspots frequented by expert birders. We found that domain-specific data augmentation is key to build models that are robust against high ambient noise levels and can cope with overlapping vocalizations. Task-specific model designs and training regimes for audio event recognition perform on-par with very complex architectures used in other domains (e.g., object detection in images). We also found that high temporal resolution of input spectrograms (short FFT window length) improves the classification performance for bird sounds. In summary, BirdNET achieved a mean average precision of 0.791 for single-species recordings, a F0.5 score of 0.414 for annotated soundscapes, and an average correlation of 0.251 with hotspot observation across 121 species and 4 years of audio data. By enabling the efficient extraction of the vocalizations of many hundreds of bird species from potentially vast amounts of audio data, BirdNET and similar tools have the potential to add tremendous value to existing and future passively collected audio datasets and may transform the field of avian ecology and conservation.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1016/j.ecoinf.2021.101236

Cites

68 of 68 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Deep learning models, such as Convolutional Neural Networks (CNN), have been developed to identify birds, bats or amphibians.

    “More recently, artificial intelligence deep learning models, such as Convolutional Neural Networks (CNN), have been developed to identify birds, bats or amphibians36–40.”

    From Soundscapes and deep learning enable tracking biodiversity recovery in tropical forests (2023) · cites this paper as [40]

  • Models pre-trained on ImageNet are configured to extract a variety of generic image features that can be useful for spectrogram recognition.

    “Models pre-trained on ImageNet are configured to extract a variety of generic image features that can be useful for spectrogram recognition40.”

    From Soundscapes and deep learning enable tracking biodiversity recovery in tropical forests (2023) · cites this paper as [40]

  • While machine learning has been extensively applied to camera trap images, its application to long-term passive acoustic monitoring datasets is still in its infancy and the first deep-learning-based studies are only starting to appear.

    “While ML has been extensively applied to camera trap images, its application to long-term PAM datasets is still in its infancy and the first DL-based studies are only starting to appear”

    From Perspectives in machine learning for wildlife conservation (2022) · cites this paper as [61]

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

Flagged first, then by citations

3 of the 71 references couldn’t be matched to a known paper.

Research works printed but not linked: 2

  1. Fitzpatrick, J.W., Rodewald, A.D., 2016. Handbook of Bird Biology, Chapter Bird Conservation. John Wiley & Sons.
  2. Ralph, C.J., Droege, S., Sauer, J.R., 1995. Monitoring Bird Populations by Point Counts, Volume 149, Chapter Managing and Monitoring Birds Using Point Counts: Standards and Applications, Pages 161-168. US Department of Agriculture, Forest Service, Pacific Southwest Research Station.

Cited by

Sorted by citations

Related

Authors
Kahl, S., Wood, C. M., Eibl, M., & Klinck, H.
Affiliations
  • Cornell University (US)
  • Chemnitz University of Technology (DE)
Corresponding author
Kahl, Stefan
ORCID
Venue
Ecological Informatics
Publisher
Elsevier BV
Published
2021-03
Volume
61
Pages
101236
ISSN
1574-9541 (ISSN-L)
DOI
10.1016/j.ecoinf.2021.101236
OpenAlex
W3121931845
Cited by
903
References
71
Citations by year
2026: 242 · 2025: 250 · 2024: 197 · 2023: 114 · 2022: 81 · 2021: 19
FWCI
47.92
Type
journal article
Language
en
Open access
Licence
Locations
Ecological Informatics (publishedVersion, cc-by)
Funders
  • Arthur Vining Davis Foundations
  • European Commission
  • European Social Fund
Topics
  • Physical Sciences › Computer Science › Signal Processing › Music and Audio Processing
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Developmental Biology › Animal Vocal Communication and Behavior
  • Physical Sciences › Computer Science › Artificial Intelligence › Speech Recognition and Synthesis
Keywords
BirdNET, sound event detection, passive acoustic monitoring, autonomous recording units, soundscape analysis, ResNet, data augmentation, mixup, ambient noise robustness, bird conservation

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review