相似性约束的深度置信网络在SAR图像目标识别的应用

特征提取是合成孔径雷达(SAR)图像目标识别的关键环节。SAR图像中存在的相干斑点和非光滑特性使得传统针对光学图像的特征提取方法变得很难应用。虽然可以采用深度置信网络(DBN)自动地进行特征学习，但是该方法属于无监督学习方法，这使得学习到的特征与具体的任务是无关的。该文提出一种叫做相似性约束的受限玻尔兹曼机模型。该模型在学习过程中通过约束特征向量之间的相似性达到引入监督信息的目的。另外，可以将多个相似性约束的受限玻尔兹曼机堆叠成一种新的深度模型，称其为相似性约束的深度置信网络模型。实验结果表明在SAR图像目标识别应用中，该方法相比主成分分析(PCA)以及原始DBN具有更好的识别性能。

Similarity Constrained Deep Belief Networks with Application to SAR Image Target Recognition

Feature extraction is a key step in SAR image target recognition. The existence of speckle and discontinuity makes the conventional methods for natural images difficult to apply. Although Deep Belief Networks (DBNs) can be used to learn feature representations automatically, they work essentially in an unsupervised way, and hence the learned features are task-irrelevant. A new Boltzmann machine called Similarity constrained Restricted Boltzmann Machines (SRBMs) is proposed, which injects the supervised information into learning process through constraint on the similarity of feature vectors. Furthermore, a deep architecture named Similarity constrained DBNs (SDBNs) is constructed by layer-wise stacking of SRBMs. Experimental results show the proposed SDBN is superior to DBN and PCA in SAR image target recognition.