Asymmetric Adaptation of Deep Features for Cross-Domain Classification in Remote Sensing Imagery
Asymmetric Adaptation of Deep Features for Cross-Domain Classification in Remote Sensing Imagery
复制标题
DOI：
10.1109/lgrs.2018.2800642
复制
发表时间：
2018-02
影响因子：
4.8
通讯作者：
N. Ammour;Laila Bashmal;Y. Bazi;Mohamad Mahmoud Al Rahhal;M. Zuair
中科院分区：
文献类型：
--
作者：
N. Ammour;Laila Bashmal;Y. Bazi;Mohamad Mahmoud Al Rahhal;M. Zuair
在这封信中，我们介绍了一种非对称自适应神经网络（AANN）方法的跨域遥感图像分类。在适应过程之前，我们将从预训练的卷积神经网络中获得的特征馈送到去噪自动编码器（DAE）以执行降维。然后，AANN的第一个隐藏层（位于DAE的顶部）将标记的源数据映射到目标空间，而后续层控制可用土地覆盖类之间的分离。为了学习其权重，网络最小化目标函数，该目标函数由与源和目标数据分布之间的距离以及类分离相关的两个损失组成。根据三个基准场景遥感数据集（即，默塞德，KSA和AID数据集）的报告和讨论。
In this letter, we introduce an asymmetric adaptation neural network (AANN) method for cross-domain classification in remote sensing images. Before the adaptation process, we feed the features obtained from a pretrained convolutional neural network to a denoising autoencoder (DAE) to perform dimensionality reduction. Then the first hidden layer of AANN (placed on the top of DAE) maps the labeled source data to the target space, while the subsequent layers control the separation between the available land-cover classes. To learn its weights, the network minimizes an objective function composed of two losses related to the distance between the source and target data distributions and class separation. The results of experiments conducted on six scenarios built from three benchmark scene remote sensing data sets (i.e., Merced, KSA, and AID data sets) are reported and discussed.