TY - GEN
T1 - Small sample learning optimization for Resnet based SAR target recognition
AU - Fu, Zhenzhen
AU - Zhang, Fan
AU - Yin, Qiang
AU - Li, Ruirui
AU - Hu, Wei
AU - Li, Wei
N1 - Publisher Copyright:
© 2018 IEEE
PY - 2018/10/31
Y1 - 2018/10/31
N2 - Deep convolutional neural network (CNN) is an important branch of deep learning. Due to its strong ability of feature extraction, CNN models have been introduced to solve the problems of synthetic aperture radar automatic target recognition (SAR-ATR). However, labeled SAR images are difficult to acquire. Therefore, how to obtain a good recognition result from a small sample dataset is what we mainly focus on. In theory, a deeper network can bring a better training result. But it also brings more difficulties to the training process, especially with limited labeled training data. The residual learning which proposed in recent years can alleviate this problem effectively. In this paper, we use a deep residual network, and introduce the dropout layer into the building block to alleviate overfitting caused by limited SAR data. In order to improve the training effect, the new loss function center loss is adopted and combined with softmax loss as the supervision signal to train the deep CNN. The experimental results show that our method can achieve the classification accuracy of 99.67% with all training data, without data augmentation or pre-training. When data of the training dataset was reduced to 20%, we can still achieve a recognition result higher than 94%.
AB - Deep convolutional neural network (CNN) is an important branch of deep learning. Due to its strong ability of feature extraction, CNN models have been introduced to solve the problems of synthetic aperture radar automatic target recognition (SAR-ATR). However, labeled SAR images are difficult to acquire. Therefore, how to obtain a good recognition result from a small sample dataset is what we mainly focus on. In theory, a deeper network can bring a better training result. But it also brings more difficulties to the training process, especially with limited labeled training data. The residual learning which proposed in recent years can alleviate this problem effectively. In this paper, we use a deep residual network, and introduce the dropout layer into the building block to alleviate overfitting caused by limited SAR data. In order to improve the training effect, the new loss function center loss is adopted and combined with softmax loss as the supervision signal to train the deep CNN. The experimental results show that our method can achieve the classification accuracy of 99.67% with all training data, without data augmentation or pre-training. When data of the training dataset was reduced to 20%, we can still achieve a recognition result higher than 94%.
KW - Automatic target recognition (ATR)
KW - Center loss
KW - Convolutional neural network (CNN)
KW - Limited labeled data
KW - Residual learning
KW - Synthetic aperture radar (SAR)
UR - http://www.scopus.com/inward/record.url?scp=85063159887&partnerID=8YFLogxK
U2 - 10.1109/IGARSS.2018.8517574
DO - 10.1109/IGARSS.2018.8517574
M3 - Conference contribution
AN - SCOPUS:85063159887
T3 - International Geoscience and Remote Sensing Symposium (IGARSS)
SP - 2330
EP - 2333
BT - 2018 IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018 - Proceedings
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 38th Annual IEEE International Geoscience and Remote Sensing Symposium, IGARSS 2018
Y2 - 22 July 2018 through 27 July 2018
ER -