Bat Detective - Deep Learning Tools for Bat Acoustic Signal Detection

Abstract

Summary1. Passive acoustic sensing has emerged as a powerful tool for quantifying anthropogenic impacts on biodiversity, especially for echolocating bat species.To better assess bat population trends there is a critical need for accurate, reliable, and open source tools that allow the detection and classification of bat calls in large collections of audio recordings. The majority of existing tools are commercial or have focused on the species classification task, neglecting the important problem of first localizing echolocation calls in audio which is particularly problematic in noisy recordings.2. We developed a convolutional neural network (CNN) based open-source pipeline for detecting ultrasonic, full-spectrum, search-phase calls produced by echolocating bats (BatDetect). Our deep learning algorithms (CNNFULL and CNNFAST) were trained on full-spectrum ultrasonic audio collected along roadtransects across Romania and Bulgaria by citizen scientists as part of the iBats programme and labelled by users of www.batdetective.org. We compared the performance of our system to other algorithms and commercial systems on expert verified test datasets recorded from different sensors and countries. As an example application, we ran our detection pipeline on iBats monitoring data collected over five years from Jersey (UK), and compared results to a widelyused commercial system.3. Here, we show that both CNNFULL and CNNFAST deep learning algorithms have a higher detection performance (average precision, and recall) of search-phase echolocation calls with our test sets, when compared to other existing algorithms peer-reviewed) is the author/funder. All rights reserved. No reuse allowed without permission.The copyright holder for this preprint (which was not . http://dx.doi.org/10.1101/156869 doi: bioRxiv preprint first posted online Jun. 29, 2017; Deep Learning Tools for Bat Acoustic Signal Detection -Mac Aodha et al. 4and commercial systems tested. Precision scores for commercial systems were reasonably good across all test datasets (>0.7), but this was at the expense of recall rates. In particular, our deep learning approaches were better at detecting calls in road-transect data, which contained more noisy recordings. Our comparison of CNNFULL and CNNFAST algorithms was favourable, although CNNFAST had a slightly poorer performance, displaying a trade-off between speed and accuracy. Our example monitoring application demonstrated that our opensource, fully automatic, BatDetect CNNFAST pipeline does as well or better compared to a commercial system with manual verification previously used to analyse monitoring data.4. We show that it is possible to both accurately and automatically detect bat search-phase echolocation calls, particularly from noisy audio recordings. Our detection pipeline enables the automatic detection and monitoring of bat populations, and further facilitates their use as indicator species on a large scale, particularly when combined with automatic species identification. We release our system and datas...

Machine learning in acoustics: Theory and applications

“…An early use of this in bioacoustics can be seen in the work of Halkias et al (2013) 191 that demonstrated the ability of deep Boltzmann machines to distinguish mysticete species. Deep CNNs have been used for bat species identification 204 and have become one of the dominant types of recognizers for bird species identification since the successful introduction of CNNs in the LifeCLEF bird identification task. 205 Unsupervised ML has not been used as extensively in bioacoustics but is nonetheless present.…”

Automatic Detection and Compression for Passive Acoustic Monitoring of the African Forest Elephant

“…Given a short audio-clip we want to classify it as containing a signal produced by the species of interest (in this case an elephant rumble) or not. This setting has been considered by many previous authors (Mac Aodha et al 2018;Nichols 2016;Bittle and Duncan 2013), and is a crucial stepping stone toward using PAM for population surveying and monitoring. A similar setting we consider is one of segmentation where we want to classify each discrete time step as belonging to an elephant call or not.…”

“…Popular strategies include SVMs based upon MFCC (Dufour et al 2014), segmentation via deep learning (Koops, Van Balen, and Wiering 2015) and dictionary learning (Salamon et al 2017). Beyond birds, insects (Ganchev and Potamitis 2007), bats (Mac Aodha et al 2018) and monkeys (Turesson et al 2016) have all been considered. Elephants have been studied from a similar perspective to ours by Pleiss, Wrege, and Gomes.…”

Separating overlapping bat calls with a bi-directional long short-term memory network

“…Convolutional neural networks (CNNs) have been used to predict the presence of a search-phase bat echolocation call in spectrograms. This binary classification problem was used to detect the presence of bats [2]. To our knowledge, the use of deep learning techniques to separate animal calls that overlap in both time and frequency space has yet to be reported.…”

“…The primary objective of this study is to develop a technique for separating two target signals (echolocation and socialization calls) from mixtures of acoustic sounds. Although deep leaning has been employed in the acoustic classification of multiple species, including nonhuman primates [28], birds [4], whales [5], and bats [2, 3], the goal of the present study is distinct from these previous cases in which deep neural networks were primarily used as classifiers.…”