TY - JOUR
T1 - Underwater target classification at greater depths using deep neural network with joint multiple-domain feature
AU - Cao, Xu
AU - Zhang, Xiaomin
AU - Togneri, Roberto
AU - Yu, Yang
PY - 2019/3/1
Y1 - 2019/3/1
N2 - For underwater target classification which is supposed to recognise different ships with the radiated acoustic signal, it is the most challenging task to provide excellent classification accuracy in a variety of environments. However, most of the existing systems are optimised to get the best performance on the data set from certain situations which they are trained in, which may lead to generalisation risks when applied to new environments. Here, the authors introduce an underwater target classification framework using a deep neural network to learn deep features from a large joint multiple-domain input. The authors propose to incorporate spectral and wavelet domain information with different resolutions to grasp the 'global' structure and the 'local' transient variation of the raw radiated signals. In contrast to shallow models, a stacked sparse autoencoder (SSAE) model, which is composed of multiple hidden layers and a softmax classifier, is adopted to learn more discriminating features for classification. In the authors' experiments, the proposed SSAE model is evaluated on the data set consisting of underwater acoustic signal received at different ocean depths. The authors' results show that the proposed SSAE model with joint input features achieved a 5% improvement in classification accuracy compared to the state-of-the-art DBN approach.
AB - For underwater target classification which is supposed to recognise different ships with the radiated acoustic signal, it is the most challenging task to provide excellent classification accuracy in a variety of environments. However, most of the existing systems are optimised to get the best performance on the data set from certain situations which they are trained in, which may lead to generalisation risks when applied to new environments. Here, the authors introduce an underwater target classification framework using a deep neural network to learn deep features from a large joint multiple-domain input. The authors propose to incorporate spectral and wavelet domain information with different resolutions to grasp the 'global' structure and the 'local' transient variation of the raw radiated signals. In contrast to shallow models, a stacked sparse autoencoder (SSAE) model, which is composed of multiple hidden layers and a softmax classifier, is adopted to learn more discriminating features for classification. In the authors' experiments, the proposed SSAE model is evaluated on the data set consisting of underwater acoustic signal received at different ocean depths. The authors' results show that the proposed SSAE model with joint input features achieved a 5% improvement in classification accuracy compared to the state-of-the-art DBN approach.
KW - feature extraction
KW - neural nets
KW - learning (artificial intelligence)
KW - signal classification
KW - wavelet transforms
KW - belief networks
KW - pattern classification
KW - acoustic signal processing
KW - image classification
KW - joint input features
KW - greater depths
KW - deep neural network
KW - multiple-domain feature
KW - different ships
KW - radiated acoustic signal
KW - excellent classification accuracy
KW - underwater target classification framework
KW - deep features
KW - joint multiple-domain input
KW - domain information
KW - different resolutions
KW - raw radiated signals
KW - stacked sparse autoencoder model
KW - multiple hidden layers
KW - discriminating features
KW - SSAE model
KW - underwater acoustic signal
KW - different ocean depths
U2 - 10.1049/iet-rsn.2018.5279
DO - 10.1049/iet-rsn.2018.5279
M3 - Article
SN - 1751-8784
VL - 13
SP - 484
EP - 491
JO - IET Radar, Sonar & Navigation
JF - IET Radar, Sonar & Navigation
IS - 3
ER -