Classification of small-scale hyperspectral images with multi-source deep transfer learning
Classification of small-scale hyperspectral images with multi-source deep transfer learning
复制标题
利用多源深度迁移学习对小尺度高光谱图像进行分类
DOI：
10.1080/2150704x.2020.1714772
复制
发表时间：
2020-04-02
影响因子：
2.3
通讯作者：
Zhu, Fei
中科院分区：
文献类型：
--
作者：
Zhao, Xin;Liang, Yi;Zhu, Fei
摘要近年来，许多基于深度学习（DL）的方法被用于高光谱图像（HSI）分类，并取得了良好的效果。但这种方法通常需要大量的标记训练样本。当在小规模HSIs上应用DL时，该问题加剧。为了缓解这个问题，迁移学习（TL）被引入到HSI分析的一些现有的工作。这些作品中的大多数将知识从单个源域转移到目标域。然而，单一来源的目标语倾向于学习特定的知识，而不是一般的知识。此外，由于样本是有限的在一个源域中，它只是部分地弥补了标记样本的短缺。为了学习更多的通用知识，进一步缓解样本有限的问题，我们引入了多源迁移学习策略来分类HSI。具体来说，提出了一个名为多源深度迁移学习（MS-DTL）的框架。该框架包括一个多源兼容模型和一个定制的损失函数。我们进行实验，通过比较所提出的方法与基线方法在知名的高光谱数据集。实验结果表明，所提出的MS-DTL在小规模HSI的分类任务上优于基准测试。由于TL的策略，所提出的网络也是节省时间的。
ABSTRACT Recently, many methods based on deep learning (DL) have been used for hyperspectral image (HSI) classification and achieved good performance. But such approaches often need numerous labelled training samples. This issue is aggravated when applying DL on small-scale HSIs. To alleviate this problem, transfer learning (TL) is introduced to HSI analysis by some existing works. Most of these works transfer knowledge from a single source domain to the target domain. However, the single source TL tends to learn specific knowledge instead of general knowledge. Moreover, since the samples are limited in one source domain, it only partially alleviates the shortage of labelled samples. To learn more general knowledge and further alleviate the issue of limited samples, we introduce the multi-source transfer learning strategy to classify HSIs. Specifically, a framework named multi-source deep transfer learning (MS-DTL) is proposed. This framework consists of a multi-source compatible model and a customized loss function. We perform experiments by comparing the proposed method with the baseline methods on the well-known hyperspectral datasets. The results show that the proposed MS-DTL performs better than the benchmarks on the classification tasks of the small-scale HSIs. Thanks to the strategy of TL, the proposed network is also time-saving.