2021
DOI: 10.1038/s41598-021-96446-w
|View full text |Cite
|
Sign up to set email alerts
|

Comparing recurrent convolutional neural networks for large scale bird species classification

Abstract: We present a deep learning approach towards the large-scale prediction and analysis of bird acoustics from 100 different bird species. We use spectrograms constructed on bird audio recordings from the Cornell Bird Challenge (CBC)2020 dataset, which includes recordings of multiple and potentially overlapping bird vocalizations with background noise. Our experiments show that a hybrid modeling approach that involves a Convolutional Neural Network (CNN) for learning the representation for a slice of the spectrogr… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

0
27
0

Year Published

2022
2022
2024
2024

Publication Types

Select...
5
3
1

Relationship

0
9

Authors

Journals

citations
Cited by 68 publications
(27 citation statements)
references
References 44 publications
0
27
0
Order By: Relevance
“…Although challenges remain with respect to scalability, computational efficiency, and how to handle depauperate data (Alom et al, 2018), deep learning is one of the most powerful analytical tools in the modern researcher's toolbox, particularly when human knowledge is lacking, or datasets are too large to be workable by traditional means. In the context of ecology and evolutionary biology, there have been many recent applications of both shallow and deep machine learning, including population genetics and phylogeography (e.g., Schrider & Kern, 2018;Provost et al, 2021), bioacoustics (e.g., Nicholson, 2016;Zhong et al, 2020;Cohen et al, bioRxiv), species classification (e.g., Gupta et al, 2021), phylogenetics (e.g., Halgaswaththa et al, 2012;Wang et al, 2016), sequencing and genomics (e.g., Adrion et al, 2020;Boža et al, 2017), and phenotypic analyses and morphometrics (e.g., Devine et al, 2020;Lürig et al, 2021). Neural networks and support vector machines tend to be the most applied algorithms towards these analyses.…”
Section: Machine Learningmentioning
confidence: 99%
See 1 more Smart Citation
“…Although challenges remain with respect to scalability, computational efficiency, and how to handle depauperate data (Alom et al, 2018), deep learning is one of the most powerful analytical tools in the modern researcher's toolbox, particularly when human knowledge is lacking, or datasets are too large to be workable by traditional means. In the context of ecology and evolutionary biology, there have been many recent applications of both shallow and deep machine learning, including population genetics and phylogeography (e.g., Schrider & Kern, 2018;Provost et al, 2021), bioacoustics (e.g., Nicholson, 2016;Zhong et al, 2020;Cohen et al, bioRxiv), species classification (e.g., Gupta et al, 2021), phylogenetics (e.g., Halgaswaththa et al, 2012;Wang et al, 2016), sequencing and genomics (e.g., Adrion et al, 2020;Boža et al, 2017), and phenotypic analyses and morphometrics (e.g., Devine et al, 2020;Lürig et al, 2021). Neural networks and support vector machines tend to be the most applied algorithms towards these analyses.…”
Section: Machine Learningmentioning
confidence: 99%
“…In the context of ecology and evolutionary biology, there have been many recent applications of both shallow and deep machine learning, including population genetics and phylogeography [e.g., 46, 47], bioacoustics [e.g., 48, 49, 50], species classification [e.g., 51], phylogenetics [e.g., 52, 53], sequencing and genomics [e.g., 54, 55], and phenotypic analyses and morphometrics [e.g., 56, 57]. Neural networks and support vector machines tend to be the most applied algorithms in these analyses.…”
Section: Introductionmentioning
confidence: 99%
“…For the last decade or more, interest in using acoustics for ecological monitoring has been steadily increasing, bolstered by a drop in the price for recording hardware and storage (Roe et al, 2021) and more recently by advances in automated analysis (Xie et al, 2019). For a number of years, deep learning techniques have dominated these automated analysis approaches (Gupta et al, 2021). In the 2018 Bird Audio Detection challenge a competition for classifying 10-s audio clips as containing a bird or not, the highest performing entries were all convolutional neural networks, with the most accurate results achieved using a transfer learning setup, with both resnet50 and inception models (Lasseck, 2018).…”
Section: Related Researchmentioning
confidence: 99%
“…Cakir et al [ 49 ] proposed a method combining CNN with RNN to realize automatic detection of bird calls and obtained an 88.5% AUC score on evaluation data. Gupta et al [ 50 ] novelly proposed a deep learning approach integrating CNN with CNN or RNN of LSTM, GRU, and LMU. The approach systematically compared the multiple hybrid models of CNN with CNN, or CNN with RNN.…”
Section: Introductionmentioning
confidence: 99%