A distributed system for capturing audio tracks for training Ae. aegypti mosquito detection ANNs
Abstract
The Aedes Vigilance project aims to enhance the surveillance of the Aedes aegypti mosquito, the vector of diseases like Zika, chikungunya, and dengue, by using low-cost devices such as smartphones to monitor its characteristic sound. A neural network (RNA) was developed to identify the mosquito based on the sound of its wingbeats, though it has only been tested in controlled environments due to the lack of appropriate datasets. This work proposes MosqMon, a distributed monitoring system designed to collect audio samples in uncontrolled environments, such as universities and residences, to generate realistic datasets for RNA training. The architecture follows an agent/manager model, where agents capture audio samples and send them to a manager, which stores them for training or automated classification. The system’s viability has been demonstrated through a functional implementation, and MosqMon is expected to provide essential data for improving RNA-based mosquito surveillance.
References
ISGlobal. Mosquito: el animal más letal del mundo. Accessed: 2024-10-17. 2023. Disponível em: [link].
Bradshaw, C. J. et al. Massive yet grossly underestimated global costs of invasive insects. Nature communications, Nature Publishing Group UK London, v. 7, n. 1, p. 12986, 2016.
Diagne, C. et al. High and rising economic costs of biological invasions worldwide. Nature, Nature Publishing Group UK London, v. 592, n. 7855, p. 571–576, 2021.
Ahmed, D. A. et al. Managing biological invasions: the cost of inaction. Biological Invasions, Springer, v. 24, n. 7, p. 1927–1946, 2022.
Saúde, M. da. Painel de Monitoramento das Arboviroses — Ministério da Saúde. 2024. [link]. Acessado em: 19 de setembro de 2024.
Panjwani, A.; Wilson, A. What is stopping the use of genetically modified insects for disease control? PLoS Pathogens, Public Library of Science San Francisco, CA USA, v. 12, n. 10, e1005830, 2016.
Legros, M. et al. Gene drive strategies of pest control in agricultural systems: Challenges and opportunities. Evolutionary applications, Wiley Online Library, v. 14, n. 9, p. 2162–2178, 2021.
Paim, K. O. et al. Acoustic identification of Ae. aegypti mosquitoes using smartphone apps and residual convolutional neural networks. Biomedical Signal Processing and Control, v. 95, p. 106342, 2024. ISSN 1746-8094. DOI: 10.1016/j.bspc.2024.106342. Disponível em: [link].
Fernandes, M. S.; Cordeiro, W.; Recamonde-Mendoza, M. Detecting Aedes aegypti mosquitoes through audio classification with convolutional neural networks. Computers in Biology and Medicine, v. 129, p. 104152, 2021. ISSN 0010-4825. DOI: 10.1016/j.compbiomed.2020.104152. Disponível em: [link].
Spitzen, J.; Takken, W. Keeping track of mosquitoes: a review of tools to track, record and analyse mosquito flight. Parasites Vectors, v. 11, p. 123, 2018. Received: 03 October 2017; Accepted: 21 February 2018; Published: 02 March 2018. DOI: 10.1186/s13071-018-2735-6.
Javed, N.; Paradkar, P. N.; Bhatti, A. An overview of technologies available to monitor behaviours of mosquitoes. Acta Tropica, v. 258, p. 107347, 2024. ISSN 0001-706X. DOI: 10.1016/j.actatropica.2024.107347. Disponível em: [link].
Lai, Z. et al. Development and evaluation of an efficient and real-time monitoring system for the vector mosquitoes, Aedes albopictus and Culex quinquefasciatus. PLOS Neglected Tropical Diseases, Public Library of Science, v. 16, n. 9, p. 1–14, set. 2022. DOI: 10.1371/journal.pntd.0010701.
Kravchenko, T.; Bogdanova, T.; Shevgunov, T. Ranking Requirements Using MoSCoW Methodology in Practice. In: Silhavy, R. (Ed.). Cybernetics Perspectives in Systems. Cham: Springer International Publishing, 2022. P. 188–199. ISBN 978-3-031-09073-8.
Rodrigues, Y. d. F.; Mansilha, R. B. MosqMon: Intelligent Mosquito Trap System. 2024. [link]. Accessed: 2024-10-21.
