期刊
REMOTE SENSING
卷 13, 期 13, 页码 -出版社
MDPI
DOI: 10.3390/rs13132575
关键词
hyperspectral image classification (HSI); deep learning; convolutional neural network; dynamic neural network; wide and deep neural network
类别
资金
- National Key R&D Program of China [2018YFC1504805]
- National Natural Science Foundation of China [61806022, 41941019, 41874005]
- Fundamental Research Funds for the Central Universities [300102269103, 300102269304, 300102260301, 300102261404, 300102120201]
- Key Research and Development Program of Shaanxi [2021NY-170]
- Special Project of Forestry Science and Technology Innovation Plan in Shaanxi Province [SXLK2021-0225]
- China Scholarship Council (CSC) [201404910404]
The paper proposed a dynamic wide and deep neural network (DWDNN) for HSI classification, including multiple efficient wide sliding window and subsampling (EWSWS) networks that can grow dynamically with problem complexity. Compared to other deep learning methods, the proposed approach achieved the highest test accuracies.
Recently, deep learning has been successfully and widely used in hyperspectral image (HSI) classification. Considering the difficulty of acquiring HSIs, there are usually a small number of pixels used as the training instances. Therefore, it is hard to fully use the advantages of deep learning networks; for example, the very deep layers with a large number of parameters lead to overfitting. This paper proposed a dynamic wide and deep neural network (DWDNN) for HSI classification, which includes multiple efficient wide sliding window and subsampling (EWSWS) networks and can grow dynamically according to the complexity of the problems. The EWSWS network in the DWDNN was designed both in the wide and deep direction with transform kernels as hidden units. These multiple layers of kernels can extract features from the low to high level, and because they are extended in the wide direction, they can learn features more steadily and smoothly. The sliding windows with the stride and subsampling were designed to reduce the dimension of the features for each layer; therefore, the computational load was reduced. Finally, all the weights were only from the fully connected layer, and the iterative least squares method was used to compute them easily. The proposed DWDNN was tested with several HSI data including the Botswana, Pavia University, and Salinas remote sensing datasets with different numbers of instances (from small to big). The experimental results showed that the proposed method had the highest test accuracies compared to both the typical machine learning methods such as support vector machine (SVM), multilayer perceptron (MLP), radial basis function (RBF), and the recently proposed deep learning methods including the 2D convolutional neural network (CNN) and the 3D CNN designed for HSI classification.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据