Seismic savanna: machine learning for classifying wildlife and behaviours using ground‐based vibration field recordings

Abstract

We develop a machine learning approach to detect and discriminate elephants from other species, and to recognise important behaviours such as running and rumbling, based only on seismic data generated by the animals. We demonstrate our approach using data acquired in the Kenyan savanna, consisting of 8000 h seismic recordings and 250 k camera trap pictures. Our classifiers, different convolutional neural networks trained on seismograms and spectrograms, achieved 80%-90% balanced accuracy in detecting elephants up to 100 m away, and over 90% balanced accuracy in recognising running and rumbling behaviours from the seismic data. We release the dataset used in this study: SeisSavanna represents a unique collection of seismic signals with the associated wildlife species and behaviour. Our results suggest that seismic data offer substantial benefits for monitoring wildlife, and we propose to further develop our methods using dense arrays that could result in a seismic shift for wildlife monitoring.

A Methodological Literature Review of Acoustic Wildlife Monitoring Using Artificial Intelligence Tools and Techniques

“…The most popular performance statistic for a machine learning model used for acoustic wildlife monitoring is accuracy, which measures the ratio of the model's corrected predictions to all of its predictions. It is a good statistic for disseminating the findings of the animals' vocalization to a wider audience since it is an easy-to-understand, basic, and intuitive means of measuring performance [27,34,40,43,44,[54][55][56][57]63,64]. However, accuracy can sometimes be a misleading metric, especially in imbalanced datasets where the number of instances of one class is significantly higher than the other.…”

Elephants and algorithms: a review of the current and future role of AI in elephant monitoring

“…Non-ML techniques have been employed to detect elephant presence from seismic data for censusing purposes, achieving 85% accuracy [154] and continuous wavelet transforms reached 90% accuracy in detecting forest elephants [155]. Within the realm of ML, elephant calls have been classified from seismic measurements using support vector machines (SVMs) with 73% accuracy [153], neural networks with 87% accuracy [156] and CNNs attaining 80–90% accuracy up to 100 m away [157].…”

Suivis acoustiques de biodiversité : perspectives et défis en milieu continental terrestre

“…Kalan et al (2016) ont utilisé la surveillance acoustique de sons à longue distance comme méthode de suivi indirect et non invasif de l'aire de répartition et des territoires de Chimpanzés (Pan troglodytes Blumenbach, 1776), vivant dans deux habitats différents (la forêt et la savane-bois). La méthode la plus courante pour localiser des animaux dans l'espace avec un réseau de plusieurs microphones placés autour du site d'étude, consiste à utiliser la différence de temps d'arrivée du son à chaque microphone pour trianguler la source du son à partir des signaux enregistrés (Suzuki et al 2018 ;Rhinehart et al 2020 ;Sumitani et al 2021) Pour répondre à ces problèmes de taille et de facilité d'utilisation, des détecteurs légers et plus faciles d'utilisation ont été développés, mais ceux-ci sont souvent malheureusement peu personnalisables. Ceux-ci ont souvent un coût initial qui peut être important et problématique pour les études à large échelle.…”