期刊
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
卷 60, 期 -, 页码 -出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TGRS.2022.3209182
关键词
Training; Three-dimensional displays; Solid modeling; Costs; Feature extraction; Convolutional neural networks; Computational modeling; 3-D convolutional neural network (3-D CNN); active learning (AL); hyperspectral image classification (HSIC); spatial-spectral information; transfer learning
类别
资金
- National Natural Science Foundation of China [42271350, 62201553]
This study introduces a fast and compact active learning method for hyperspectral image classification using 3-D CNN, which integrates deep transfer learning and active learning, achieving higher accuracy with very few training samples. By querying the most informative and heterogeneous samples from the validation set, the computational cost is reduced significantly.
Convolutional neural networks (CNNs) have been extensively studied for hyperspectral image classification (HSIC). However, CNNs are critically attributed to a large number of labeled training samples, which outlays high costs in terms of time and resources. Moreover, CNNs are trained on some samples and have been tested on the entire HSI. Perhaps, the entire HSI is taken into account at test time to appropriately generate the ground-truth maps. To obtain a higher accuracy while considering the limited availability of training samples and disjoint validation and test samples, this work proposes a fast and compact 3-D CNN-based active learning (AL) for HSIC that integrates both deep transfer learning and AL into a unified framework. In the proposed methodology, a 3-D CNN model is trained with very few training samples (i.e., 5%, only) and in the next phase, the most informative and heterogeneous samples are queried from the validation set (candidate set) based on the fuzziness, mutual information, and breaking ties of the trained model. The 3-D CNN model is later fine-tuned (rather than retraining from scratch) with the new training samples (i.e., 200 samples are selected in each iteration) to reduce the computational cost. The proposed method has been compared with the state-of-the-art traditional and deep models proposed for HSIC. Experimental results proved the superiority of our proposed method on several benchmark HSI datasets with significantly fewer labeled samples. MATLAB demo can be accessed on GitHub: github.com/mahmad00
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