Using deep learning for acoustic event classification: The case of natural disasters

Abstract

This study proposes a sound classification model for natural disasters. Deep learning techniques, a convolutional neural network (CNN) and long short-term memory (LSTM), were used to train two individual classifiers. The study was conducted using a dataset acquired online1 and truncated at 0.1 s to obtain a total of 12 937 sound segments. The result indicated that acoustic signals are effective for classifying natural disasters using machine learning techniques. The classifiers serve as an alternative effective approach to disaster classification. The CNN model obtained a classification accuracy of 99.96%, whereas the LSTM obtained an accuracy of 99.90%. The misclassification rates obtained in this study for the CNN and LSTM classifiers (i.e., 0.4% and 0.1%, respectively) suggest less classification errors when compared to existing studies. Future studies may investigate how to implement such classifiers for the early detection of natural disasters in real time.

Underwater Noise Modeling and Its Application in Noise Classification with Small-Sized Samples

“…Li et al [27] developed a U-Net model based on a random mode-coupling matrix to recover distorted acoustic interference striations. Ekpez proposed a method for natural disasters classification, and the CNN model obtained a classification accuracy of 99.96%, whereas the LSTM obtained an accuracy of 99.90% [28]. Merchant ship-radiated noise was employed for seabed classification using an ensemble of deep learning (DL) algorithms by Escobar-Amado.…”

Acoustic scene classification based on three-dimensional multi-channel feature-correlated deep learning networks

“…With the availability of large-scale datasets due to the vast amount of audio clips recorded by mobile devices, statistical models are gradually replaced by deep-learning methods, which are generally based on convolution neural networks (CNN) and are shown to achieve much better performance in the ASC task. In 15 , 16 , the authors summarized the performance of various deep-learning backbone models as typically used in computer visions, e.g., VGG 17 , Xception 18 , ResNet 19 , recurrent neural networks (RNNs) 20 , and long-short-term memory networks (LSTM) 21 , and confirmed that performance superiorities over statistical approaches can be well generalized to acoustic scenes. Moreover, it is shown in 22 – 24 that the so-called ensembled-learning approach based on the aggregation of several CNN models effectively improves performance as compared with a single deep-learning model due to the diversities of extracted features and system architectures 25 , 26 .…”