Enhanced-Random-Feature-Subspace-Based Ensemble CNN for the Imbalanced Hyperspectral Image Classification

Hyperspectral image (HSI) classification often faces the problem of multiclass imbalance, which is considered to be one of the major challenges in the field of remote sensing. In recent years, deep learning has been successfully applied to the HSI classification, a convolutional neural network (CNN) is one of the most representative of them. However, it is difficult to effectively improve the accuracy of minority classes under the problem of multiclass imbalance. In addition, ensemble learning has been successfully applied to solve multiclass imbalance, such as random forest (RF) This article proposes a novel enhanced-random-feature-subspace-based ensemble CNN algorithm for the multiclass imbalanced problem. The main idea is to perform random oversampling of training samples and multiple data enhancements based on random feature subspace, and then, construct an ensemble learning model combining random feature selection and CNN to the HSI classification. Experimental results on three public hyperspectral datasets show that the performance of the proposed method is better than the traditional CNN, RF, and deep learning ensemble methods.

[1]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[2]  C. Lee Giles,et al.  Learning on the border: active learning in imbalanced data classification , 2007, CIKM '07.

[3]  Xuesong Wang,et al.  Spectral–Spatial Feature Extraction for HSI Classification Based on Supervised Hypergraph and Sample Expanded CNN , 2018, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[4]  Ying Wang,et al.  Deep Learning Ensemble for Hyperspectral Image Classification , 2019, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[5]  Angshul Majumdar,et al.  Discriminative Robust Deep Dictionary Learning for Hyperspectral Image Classification , 2017, IEEE Transactions on Geoscience and Remote Sensing.

[6]  Bartosz Krawczyk,et al.  Analyzing the oversampling of different classes and types of examples in multi-class imbalanced datasets , 2016, Pattern Recognit..

[7]  Xiuping Jia,et al.  Hyperspectral Image Classification Using Convolutional Neural Networks and Multiple Feature Learning , 2018, Remote. Sens..

[8]  Wei Feng,et al.  New margin-based subsampling iterative technique in modified random forests for classification , 2019, Knowl. Based Syst..

[9]  Jinchang Ren,et al.  ANN vs. SVM: Which one performs better in classification of MCCs in mammogram imaging , 2012, Knowl. Based Syst..

[10]  Fang Liu,et al.  Imbalanced Hyperspectral Image Classification Based on Maximum Margin , 2015, IEEE Geoscience and Remote Sensing Letters.

[11]  Leo Breiman,et al.  Bagging Predictors , 1996, Machine Learning.

[12]  Andrew Zisserman,et al.  Very Deep Convolutional Networks for Large-Scale Image Recognition , 2014, ICLR.

[13]  Ying Li,et al.  Spectral-Spatial Classification of Hyperspectral Imagery with 3D Convolutional Neural Network , 2017, Remote. Sens..

[14]  Sandra Lowe,et al.  Classification Methods For Remotely Sensed Data , 2016 .

[15]  Jon Atli Benediktsson,et al.  Deep Learning for Hyperspectral Image Classification: An Overview , 2019, IEEE Transactions on Geoscience and Remote Sensing.

[16]  David H. Wolpert,et al.  Stacked generalization , 1992, Neural Networks.

[17]  Wei Feng,et al.  Imbalanced Hyperspectral Image Classification With an Adaptive Ensemble Method Based on SMOTE and Rotation Forest With Differentiated Sampling Rates , 2019, IEEE Geoscience and Remote Sensing Letters.

[18]  Jon Atli Benediktsson,et al.  Hyperspectral Image Classification With Rotation Random Forest Via KPCA , 2017, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[19]  Peijun Du,et al.  Random Forest and Rotation Forest for fully polarized SAR image classification using polarimetric and spatial features , 2015 .

[20]  Paul M. Mather,et al.  Classification methods for remotely sensed data, 2nd ed , 2016 .

[21]  Yuanchao Su,et al.  Random Subspace Ensemble With Enhanced Feature for Hyperspectral Image Classification , 2020, IEEE Geoscience and Remote Sensing Letters.

[22]  Mingquan Wu,et al.  Dynamic Synthetic Minority Over-Sampling Technique-Based Rotation Forest for the Classification of Imbalanced Hyperspectral Data , 2019, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[23]  Chee Khiang Pang,et al.  Kernel-based SMOTE for SVM classification of imbalanced datasets , 2015, IECON 2015 - 41st Annual Conference of the IEEE Industrial Electronics Society.

[24]  Fan Zhang,et al.  Deep Convolutional Neural Networks for Hyperspectral Image Classification , 2015, J. Sensors.

[25]  Wei Li,et al.  Hyperspectral Image Classification With Imbalanced Data Based on Orthogonal Complement Subspace Projection , 2018, IEEE Transactions on Geoscience and Remote Sensing.

[26]  Nicholas C. Coops,et al.  Towards the Operational Use of Satellite Hyperspectral Image Data for Mapping Nutrient Status and Fertilizer Requirements in Australian Plantation Forests , 2013, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[27]  Wei Feng,et al.  Weight-Based Rotation Forest for Hyperspectral Image Classification , 2017, IEEE Geoscience and Remote Sensing Letters.

[28]  Laurie A. Chisholm,et al.  Classification of Australian Native Forest Species Using Hyperspectral Remote Sensing and Machine-Learning Classification Algorithms , 2014, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[29]  Pabitra Mitra,et al.  BASS Net: Band-Adaptive Spectral-Spatial Feature Learning Neural Network for Hyperspectral Image Classification , 2016, IEEE Transactions on Geoscience and Remote Sensing.

[30]  R. Schapire.  The Strength of Weak Learnability , 1990, Machine Learning.

[31]  Paolo Gamba,et al.  A novel framework for urban mapping from multispectral and hyperspectral data , 2013 .

[32]  Somsak Choomchuay,et al.  Comparison of Sampling Methods for Imbalanced Data Classification in Random Forest , 2018, 2018 11th Biomedical Engineering International Conference (BMEiCON).

[33]  Zhiping Lin,et al.  Kernel based online learning for imbalance multiclass classification , 2018, Neurocomputing.

[34]  Yoshua Bengio,et al.  Learning long-term dependencies with gradient descent is difficult , 1994, IEEE Trans. Neural Networks.

[35]  Peijun Du,et al.  Spectral–Spatial Classification for Hyperspectral Data Using Rotation Forests With Local Feature Extraction and Markov Random Fields , 2015, IEEE Transactions on Geoscience and Remote Sensing.

[36]  Dumitru Erhan,et al.  Going deeper with convolutions , 2014, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[37]  Blockin.  SVM : Which One Performs Better in Classification of MCCs in Mammogram Imaging , 2022 .

[38]  Peijun Du,et al.  Random Subspace Ensembles for Hyperspectral Image Classification With Extended Morphological Attribute Profiles , 2015, IEEE Transactions on Geoscience and Remote Sensing.

[39]  Seetha Hari,et al.  Learning From Imbalanced Data , 2019, Advances in Computer and Electrical Engineering.

[40]  Francisco Herrera,et al.  Dynamic ensemble selection for multi-class imbalanced datasets , 2018, Inf. Sci..

[41]  Francisco Herrera,et al.  A Review on Ensembles for the Class Imbalance Problem: Bagging-, Boosting-, and Hybrid-Based Approaches , 2012, IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews).

[42]  Lin Zhu,et al.  Hyperspectral Images Classification With Gabor Filtering and Convolutional Neural Network , 2017, IEEE Geoscience and Remote Sensing Letters.

[43]  Øystein Haugen,et al.  Boosting algorithms for network intrusion detection: A comparative evaluation of Real AdaBoost, Gentle AdaBoost and Modest AdaBoost , 2020, Eng. Appl. Artif. Intell..