Litcius/Paper detail

Multiscale Deep Neural Network With Two-Stage Loss for SAR Target Recognition With Small Training Set

Jian Guan, Jiabei Liu, Pengming Feng, Wenwu Wang

2021IEEE Geoscience and Remote Sensing Letters26 citationsDOI

Abstract

Deep learning models have been used recently for target recognition from synthetic aperture radar (SAR) images. However, the performance of these models tends to deteriorate when only a small number of training samples are available due to the problem of overfitting. To address this problem, we propose a two-stage multiscale densely connected convolutional neural networks (TMDC-CNNs). In the proposed TMDC-CNNs, the overfitting issue is addressed with a novel multiscale densely connected network architecture and a two-stage loss function, which integrated the cosine similarity with the prevailing softmax cross-entropy loss. Experiments were conducted on the MSTAR data set, and the results show that our model offers significant recognition accuracy improvements as compared with other state-of-the-art methods, with severely limited training data. The source codes are available at <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/Stubsx/TMDC-CNNs</uri> .

Topics & Concepts

OverfittingSoftmax functionArtificial intelligenceComputer scienceSynthetic aperture radarPattern recognition (psychology)Convolutional neural networkDeep learningAutomatic target recognitionArtificial neural networkMachine learningAdvanced SAR Imaging TechniquesSynthetic Aperture Radar (SAR) Applications and TechniquesGeophysical Methods and Applications
Multiscale Deep Neural Network With Two-Stage Loss for SAR Target Recognition With Small Training Set | Litcius