Utilizing contrastive learning for graph-based active learning of SAR data
Utilizing contrastive learning for graph-based active learning of SAR data
复制标题
DOI：
10.1117/12.2663099
复制
发表时间：
2023-06
期刊：
影响因子：
--
通讯作者：
Jason S. Brown;Riley O'Neill;J. Calder;A. Bertozzi
中科院分区：
文献类型：
--
作者：
Jason S. Brown;Riley O'Neill;J. Calder;A. Bertozzi
利用合成孔径雷达（SAR）数据进行自动目标识别是一个具有挑战性的图像分类问题，因为传统深度学习方法很难获得大量标记训练集。最近的工作1通过利用基于图的半监督学习和主动学习中的强大工具来解决这个问题，并在移动和静止目标获取和识别（MSTAR）数据集上取得了最先进的结果，与现有技术相比，标记数据较少。之前工作的一个关键部分是使用无监督深度学习，特别是卷积变分自动编码器，在构建相似性图之前将MSTAR图像嵌入到有意义的特征空间中。在本文中，我们开发了一个对比的Simplified框架，通过使用特定的SAR图像的数据增强从MSTAR图像的特征提取。我们表明，我们的对比嵌入结果在MSTAR数据集上的自动目标识别中的变分自动编码器相似性图方法的性能得到了提高。我们还通过以不同的标签率训练支持向量机（SVM）、应用谱聚类和评估图切割能量来对自动编码器和对比嵌入的质量进行比较研究，所有这些都表明对比学习嵌入优于自动编码器嵌入的上级。
Automatic target recognition with synthetic aperture radar (SAR) data is a challenging image classification problem due to the difficulty in acquiring the large labeled training sets required for conventional deep learning methods. Recent work1 addressed this problem by utilizing powerful tools in graph-based semi-supervised learning and active learning, and achieved state of the art results on the Moving and Stationary Target Acquisition and Recognition (MSTAR) dataset with less labeled data compared to existing techniques. A key part of the previous work was the use of unsupervised deep learning, in particular, a convolutional variational autoencoder, to embed the MSTAR images into a meaningful feature space prior to constructing the similarity graph. In this paper, we develop a contrastive SimCLR framework for feature extraction from MSTAR images by using data augmentations specific to SAR imagery. We show that our contrastive embedding results in improved performance over the variational autoencoder similarity graph method in automatic target recognition on the MSTAR dataset. We also perform a comparative study of the quality of the autoencoder and contrastive embeddings by training support vector machines (SVM) at various label rates, applying spectral clustering, and evaluating graph-cut energies, all of which show that the contrastive learning embedding is superior to the autoencoder embedding.