Semi-Supervised Deep Learning Classification for Hyperspectral Image Based on Dual-Strategy Sample Selection

Ying Li,Haokui Zhang,Jonathan Chan,Bei Fang

doi:10.3390/rs10040574

Ying Li, Haokui Zhang + Show 2 more

Open Access

https://doi.org/10.3390/rs10040574

Copy DOI

Abstract

This paper studies the classification problem of hyperspectral image (HSI). Inspired by the great success of deep neural networks in Artificial Intelligence (AI), researchers have proposed different deep learning based algorithms to improve the performance of hyperspectral classification. However, deep learning based algorithms always require a large-scale annotated dataset to provide sufficient training. To address this problem, we propose a semi-supervised deep learning framework based on the residual networks (ResNets), which use very limited labeled data supplemented by abundant unlabeled data. The core of our framework is a novel dual-strategy sample selection co-training algorithm, which can successfully guide ResNets to learn from the unlabeled data by making full use of the complementary cues of the spectral and spatial features in HSI classification. Experiments on the benchmark HSI dataset and real HSI dataset demonstrate that, with a small number of training data, our approach achieves competitive performance with maximum improvement of 41% (compare with traditional convolutional neural network (CNN) with 5 initial training samples per class on Indian Pines dataset) for HSI classification as compared with the results from those state-of-the-art supervised and semi-supervised methods.

Highlights

Hyperspectral image (HSI) collected by imaging spectrometers captured rich spectral and spatial information simultaneously [1]
The effectiveness of the proposed method is tested in the classification of three open source hyperspectral datasets, namely, the Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) Indian Pines datasets, the Reflective Optics System Imaging Spectrometer (ROSIS-03) University of Pavia datasets, the AVIRIS Salinas Valley datasets and one selected Hyperion dataset
The performance of the proposed method is compared with three state-of-the-art HSI classification methods: (1) convolutional neural networks (CNNs) [10], a supervised classification using deep CNN to extract the joint spectral–spatial features from HSI; (2) CDL-MD-L [15], a self-training semi-supervised classification approach based on contextual deep learning classification (CDL) and multi-decision labeling (MD-L); (3) Co-DC-CNN, DC-CNN [30] is a dual-channel CNN with non-residual networks from our previous work, we extend it to a co-training approach denoted as Co-DC-CNN

Summary

Introduction

Hyperspectral image (HSI) collected by imaging spectrometers captured rich spectral and spatial information simultaneously [1]. Inspired by the great success of deep learning for image classification, remarkable efforts have been invested for spectral-spatial HSI classification by deep learning techniques in the last few years [7,8,9,10,11,12]. As described by Yue et al [9], a 2D-CNN with two convolutional layers and three subsampling layer is used to extract deep features from HSI with a spatial size of 42 × 42. Chen et al [10] used deep CNN with three convolutional layers and one fully connected layer to extract deep features from HSI with a spatial size of 27 × 27. Li et al [11] proposed a 3D-CNN framework to extract deep spectral-spatial combined features with two 3D convolution layers, one fully connected layer and one classification layer; the spatial size is empirically set to 5 × 5

Methods

Results

Discussion

Conclusion