Search

Recently, hyperspectral image (HSI) classification has become a hot topic in the geographical images research area. Sufficient samples are required for image classes to properly train classification models. However, a class imbalance problem has emerged in hyperspectral image (HSI) datasets as some classes do not have enough samples for training, and some classes have many samples. Therefore, the performance of classifiers is likely to be biased toward the classes with the largest samples, and this can lead to a decrease in the classification accuracy. Therefore, a new deep-learning-based model is proposed for hyperspectral images generation and classification of imbalanced data. Firstly, the spectral features are extracted by a 1D convolutional neural network, whereas a 2D convolutional neural network extracts the spatial features and the extracted spatial features and spectral features are catenated into a stacked spatial–spectral feature vector. Secondly, an autoencoder model was developed to generate synthetic images for minority classes, and the image samples were balanced. The GAN model is applied to determine the synthetic images from the real ones and then enhancing the classification performance. Finally, the balanced datasets are fed to a 2D CNN model for performing classification and validating the efficiency of the proposed model. Our model and the state-of-the-art classifiers are evaluated by four open-access HSI datasets. The results showed that the proposed approach can generate better quality samples for rebalancing datasets, which in turn noticeably enhances the classification performance compared to the existing classification models. [ABSTRACT FROM AUTHOR]

Identifying dangerous events from driving behavior data has become a vital challenge in intelligent transportation systems. In this study, we compared machine and deep learning-based methods for classifying the risk levels of near-crashes. A dataset was built for the study by considering variables related to naturalistic driving, temporal data, participants, and road geometry, among others. Hierarchical clustering was applied to categorize the near-crashes into several risk levels based on high-risk driving variables. The adaptive lasso variable model was adopted to reduce factors and select significant driving risk factors. In addition, several machine and deep learning models were used to compare near-crash classification performance by training the models and examining the model with testing data. The results showed that the deep learning models outperformed the machine learning and statistical models in terms of classification performance. The LSTM model achieved the highest performance in terms of all evaluation metrics compared with the state-of-the-art models (accuracy = 96%, recall = 0.93, precision = 0.88, and F1-measure = 0.91). The LSTM model can improve the classification accuracy and prediction of most near-crash events and reduce false near-crash classification. The finding of this study can benefit transportation safety in predicting and classifying driving risk. It can provide useful suggestions for reducing the incidence of critical events and forward road crashes. [ABSTRACT FROM AUTHOR]