Bioacoustics Monitoring of Wildlife using Artificial Intelligence: A Methodological Literature Review

Abstract

Artificial intelligence (AI) is a broad computing science that has attracted significant attention in the ecological sector because of its problem-solving, deciding, and pattern recognition capabilities. Because of the large number of datasets available across spatiotemporal scales that may be used for machine learning and interpretation, bioacoustics wildlife monitoring is essential in the performance of AI techniques. Although several studies have enforced AI algorithms into the wildlife ecology, the future of this developing method in wildlife acoustic monitoring is unknown. In this study, we performed a scientific literature review covering 20 papers from 2015 and March 2022 to evaluate its application and advise future demands. During this time, we observed a considerable increase in the use of AI approaches in wildlife acoustic monitoring. Overall, bird species $(\mathbf{N}=\mathbf{12})$ received the most attention, followed by amphibians $(\mathbf{N}=\mathbf{5})$ and mammals $(\mathbf{N}=\mathbf{3})$), even though their operations are diversifying. Among the AI learnings used in bioacoustics wildlife monitoring, a convolutional neural network was highly accurate in terms of performance, had more advantages, and was replicated in multiple articles than other classification methods. Reviewing previously used AI algorithms in bioacoustics research is expected to aid in understanding the trends and identifying gaps in automatic wildlife monitoring.