Automated classification of bat echolocation call recordings with artificial intelligence

Tabak, Michael A.; Murray, Kevin L.; Reed, Ashley M.; Lombardi, J.A.; Bay, Kimberly

doi:10.1016/j.ecoinf.2021.101526

Cited by 18 publications

(18 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We achieved correct identification rates of ≥90% for 20 of these classes, including (Allen et al, 2021;Gray et al, 2019;Norouzzadeh et al, 2018). In particular, multiple studies have successfully applied CNNs to bat echolocation calls to detect and/or classify species (Chen et al, 2020;Mac Aodha et al, 2018;Pettersson, 2020;Tabak et al, 2022). However, these foundational prior efforts fell short of meeting NABat's goals to assess status and trends of bat populations well into the future.…”

Section: Discussionmentioning

confidence: 99%

“…CNNs are increasingly developed to analyse data collected through PAM techniques, camera traps, and other autonomous monitoring devices because of their proven effectiveness at automated feature extraction, image decomposition and information gathering from huge volumes of data (Allen et al, 2021; Gray et al, 2019; Norouzzadeh et al, 2018). In particular, multiple studies have successfully applied CNNs to bat echolocation calls to detect and/or classify species (Chen et al, 2020; Mac Aodha et al, 2018; Pettersson, 2020; Tabak et al, 2022). However, these foundational prior efforts fell short of meeting NABat's goals to assess status and trends of bat populations well into the future.…”

Section: Discussionmentioning

confidence: 99%

“…BatNet (Chen et al, 2020) classifies species with 86% overall accuracy, but is limited to Asian bats and requires manual extraction of signals from audio files. Tabak et al's (2022) CNN‐based model classifies 10 North American species of bats with 90% overall accuracy, but is not compatible with cloud environments like the NABat database. Additionally, this model was trained on acoustic data recorded in zero‐crossing (ZC) format, which makes the model rely on a reduced set of information that could be critical for bat call identification.…”

Section: Discussionmentioning

confidence: 99%

“…These packages also do not incorporate recent advances in deep learning techniques, which have enabled the application of convolutional neural networks (CNNs; Albawi et al, 2017) to bat signal identification (e.g. BatDetective (Mac Aodha et al, 2018); BatNet (Chen et al, 2020); Pettersson (2020); and Tabak et al (2022)).…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

NABat ML: Utilizing deep learning to enable crowdsourced development of automated, scalable solutions for documenting North American bat populations

Khalighifar

Gotthold

Adams

et al. 2022

Journal of Applied Ecology

View full text Add to dashboard Cite

Bats play crucial ecological roles and provide valuable ecosystem services, yet many populations face serious threats from various ecological disturbances. The North American Bat Monitoring Program (NABat) aims to use its technology infrastructure to assess status and trends of bat populations, while developing innovative and community‐driven conservation solutions. Here, we present NABat ML, an automated machine‐learning algorithm that improves the scalability and scientific transparency of NABat acoustic monitoring. This model combines signal processing techniques and convolutional neural networks (CNNs) to detect and classify recorded bat echolocation calls. We developed our CNN model with internet‐based computing resources (‘cloud environment’), and trained it on >600,000 spectrogram images. We also incorporated species range maps to improve the robustness and accuracy of the model for future ‘unseen’ data. We evaluated model performance using a comprehensive, independent, holdout dataset. NABat ML successfully distinguished 31 classes (30 species and a noise class) with overall weighted‐average accuracy and precision rates of 92%, and ≥90% classification accuracy for 19 of the bat species. Using a single cloud‐environment computing instance, the entire model training process took <16 h. Synthesis and applications. Our convolutional neural network (CNN)‐based model, NABat ML, classifies 30 North American bat species using their recorded echolocation calls with an overall accuracy of 92%. In addition to providing highly accurate species‐level classification, NABat ML and its outputs are compatible with Bayesian and other statistical techniques for measuring uncertainty in classification. Our model is open‐source and reproducible, enabling future implementations as software on end‐user devices and cloud‐based web applications. These qualities make NABat ML highly suitable for applications ranging from grassroots community science initiatives to big‐data methods developed and implemented by researchers and professional practitioners. We believe the transparency and accessibility of NABat ML will encourage broad‐scale participation in bat monitoring, and enable development of innovative solutions needed to conserve North American bat species.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

NABat ML: Utilizing deep learning to enable crowdsourced development of automated, scalable solutions for documenting North American bat populations

Khalighifar

Gotthold

Adams

et al. 2022

Journal of Applied Ecology

View full text Add to dashboard Cite

show abstract

“…The most standard approach for such a judgment is by taking the so-called 'elbow point,' which is virtually the point where a severely decreasing or increasing curve begins to turn 'flat enough' (12,13,(16)(17)(18). Thus, this study considered the function of the rank factorization curve and used the function uik() from the R package inflection to select the optimal rank (17,19,20).…”

Section: 3proposed Uik Methodsmentioning

confidence: 99%

The Decision of the Optimal Rank of a Non-negative Matrix Factorization Model for Gene Expression Datasets Utilizing Unit Invariant Knee Method

Güven

2022

Preprint

View full text Add to dashboard Cite

BackgroundThere is a great need to develop a computational approach to analyze and exploit the information contained in gene expression data. Recent utilization of non-negative matrix factorization (NMF) in computational biology has served its capability to derive essential details from a high amount of data in particular gene expression microarrays.ObjectiveA common problem in NMF is finding the proper number rank (r) of factors. Thus, various techniques have been suggested to select the optimal value of rank factorization (r).MethodThis study focused on the unit invariant knee (UIK) method to calculate factorization rank (basis vector) of the non-negative matrix factorization (NMF) of gene expression data sets is employed. Because the UIK method requires an extremum distance estimator (EDE) that is eventually employed for inflection and identification of a knee point, this study finds the first inflection point of curvature of RSS of the proposed algorithms using the UIK method on gene expression datasets as a target matrix.ResultsComputation was conducted for the UIK task using the esGolub data set of R studio, and consequently, the distinct results of NMF was subjected to compare on different algorithms. The proposed UIK method is easy to perform, free of a priori rank value input, and does not require initial parameters that significantly influence the model’s functionality.ConclusionThis study demonstrates that the UIK method provides a credible prediction for both gene expression data and precisely estimating of simulated mutational processes data with known dimensions.

show abstract

BatNet: a deep learning‐based tool for automated bat species identification from camera trap images

Krivek,

Gillert,

Harder

et al. 2023

Remote Sens Ecol Conserv

View full text Add to dashboard Cite

Automated monitoring technologies can increase the efficiency of ecological data collection and support data‐driven conservation. Camera traps coupled with infrared light barriers can be used to monitor temperate‐zone bat assemblages at underground hibernacula, where thousands of individuals of multiple species can aggregate in winter. However, the broad‐scale adoption of such photo‐monitoring techniques is limited by the time‐consuming bottleneck of manual image processing. Here, we present BatNet, an open‐source, deep learning‐based tool for automated identification of 13 European bat species from camera trap images. BatNet includes a user‐friendly graphical interface, where it can be retrained to identify new bat species or to create site‐specific models to improve detection accuracy at new sites. Model accuracy was evaluated on images from both trained and untrained sites, and in an ecological context, where community‐ and species‐level metrics (species diversity, relative abundance, and species‐level activity patterns) were compared between human experts and BatNet. At trained sites, model performance was high across all species (F1‐score: 0.98–1). At untrained sites, overall classification accuracy remained high (96.7–98.2%), when camera placement was comparable to the training images (<3 m from the entrance; <45° angle relative to the opening). For atypical camera placements (>3 m or >45° angle), retraining the detector model with 500 site‐specific annotations achieved an accuracy of over 95% at all sites. In the ecological case study, all investigated metrics were nearly identical between human experts and BatNet. Finally, we exemplify the ability to retrain BatNet to identify a new bat species, achieving an F1‐score of 0.99 while maintaining high classification accuracy for all original species. BatNet can be implemented directly to scale up the deployment of camera traps in Europe and enhance bat population monitoring. Moreover, the pretrained model can serve as a baseline for transfer learning to automatize the image‐based identification of bat species worldwide.

show abstract

Automated classification of bat echolocation call recordings with artificial intelligence

Cited by 18 publications

References 27 publications

NABat ML: Utilizing deep learning to enable crowdsourced development of automated, scalable solutions for documenting North American bat populations

NABat ML: Utilizing deep learning to enable crowdsourced development of automated, scalable solutions for documenting North American bat populations

The Decision of the Optimal Rank of a Non-negative Matrix Factorization Model for Gene Expression Datasets Utilizing Unit Invariant Knee Method

BatNet: a deep learning‐based tool for automated bat species identification from camera trap images

Contact Info

Product

Resources

About