Performance evaluation of state-of-the art neural networks for automatic radar target classification task
Keywords:
Convolutional neural network, Radar target classification, Transfer learning, Wavelet transformAbstract
Currently, there are many modern convolutional neural networks (CNN) used in computer vision and natural language processing. However, the performance evaluation of those neural networks for radar target recognition is still room for research. Therefore, in this paper we uses transform learning method to conduct evaluating the performance of some deep learning neural networks based on Wavelet analysis dataset of the response radar signal from four items, including cylinders, discs, spheres and struncones). Based on the evalution of networks, a new model that combine EfficientNet and GoogleNet (named Efficient-GoogleNet), was proposed to improve the radar target classification accuracy. As a result, the Efficient-GoogleNet model outperforms other ones in terms of classification accuracy. Thus, it can be seen that building a suitable neural network has promising potential for automatic radar targets classification tasks.
References
S. Kazemi, B. Yonel and B. Yazici, "Deep Learning for Direct Automatic Target Recognition from SAR Data," 2019 IEEE Radar Conference (RadarConf), 2019, pp. 1-6, doi: 10.1109/RADAR.2019.8835492.
L. Wang, J. Tang and Q. Liao, "A Study on Radar Target Detection Based on Deep Neural Networks," in IEEE Sensors Letters, vol. 3, no. 3, pp. 1-4, March 2019, Art no. 7000504, doi: 10.1109/LSENS.2019.2896072.
H. Le et al., "Micro-Doppler-Radar-Based UAV Detection Using Inception-Residual Neural Network," 2020 International Conference on Advanced Technologies for Communications (ATC), 2020, pp. 177-181, doi: 10.1109/ATC50776.2020.9255454.
Y. Li, Y. Rui, Y. Gao, and J. Gao "Ground surveillance radar target classification based on 2D CNN", Proc. SPIE 11719, Twelfth International Conference on Signal Processing Systems, 117190N (20 January 2021); https://doi.org/10.1117/12.2581289.
N. Su, X. Chen, J. Guan, X. Mou, N. Liu, “Detection and Classification of Maritime Target with Micro-motion Based on CNNs,” Journal of Radars, 2018, 7(5): 565-574. doi: 10.12000/JR18077.
S. Samaras et al., “UAV Classification with Deep Learning Using Surveillance Radar Data,” Computer Vision Systems. ICVS 2019. Lecture Notes in Computer Science, vol 11754. Springer, Cham, 2019. https://doi.org/10.1007/978-3-030-34995-0_68.
D.V. Khablov, "The Direct Continuous Wavelet Transformation in the Processing of Radio-Wave Sensor Signals". Meas Tech 56, 510–516 (2013). https://doi.org/10.1007/s11018-013-0236-8.
M. Tan and Q. V. Le, “EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks,” International Conference on Machine Learning, 2019, 2019-05-28.
C. Szegedy et al., “Going Deeper with Convolutions,” 2014-09-17.
F. N. Iandola, S. Han, M. W. Moskewicz, K. Ashraf, W. J. Dally, and K. Keutzer, “SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size,” 2016-02-24.
K. He, X. Zhang, S. Ren, and J. Sun, "Deep Residual Learning for Image Recognition", arXiv:1512.03385, 2015.
M. Sandler, A. Howard, M. Zhu, A. Zhmoginov, L.-C. Chen, "MobileNetV2: Inverted Residuals and Linear Bottlenecks", arXiv:1801.04381, 2018.
