A Convolutional Neural Networks Approach to Audio Classification for Rainfall Estimation
Roberta Avanzato, Francesco Beritelli, Francesco Di Franco, Valerio Francesco Puglisi · 2019
The recent climatic changes imply an increasing manifestation of calamitous phenomena related to important hydrogeological disruptions in many parts of the earth. For this reason, an accurate estimate of rainfall levels becomes essential to be able to warn of the imminent occurrence of a calamitous event and reduce the risk to human beings. This paper proposes an approach based on Convolutional Neural Networks (CNN) to the classification of the audio signal coming from a new rainfall system.