Shuangting Wang

Bio: Shuangting Wang is an academic researcher. The author has contributed to research in topics: Computer science & Glacier. The author has an hindex of 1, co-authored 1 publications receiving 22 citations.

Learning Deep Hierarchical Spatial-Spectral Features for Hyperspectral Image Classification Based on Residual 3D-2D CNN.

Abstract: Every pixel in a hyperspectral image contains detailed spectral information in hundreds of narrow bands captured by hyperspectral sensors. Pixel-wise classification of a hyperspectral image is the cornerstone of various hyperspectral applications. Nowadays, deep learning models represented by the convolutional neural network (CNN) provides an ideal solution for feature extraction, and has made remarkable achievements in supervised hyperspectral classification. However, hyperspectral image annotation is time-consuming and laborious, and available training data is usually limited. Due to the "small-sample problem", CNN-based hyperspectral classification is still challenging. Focused on the limited sample-based hyperspectral classification, we designed an 11-layer CNN model called R-HybridSN (Residual-HybridSN) from the perspective of network optimization. With an organic combination of 3D-2D-CNN, residual learning, and depth-separable convolutions, R-HybridSN can better learn deep hierarchical spatial-spectral features with very few training data. The performance of R-HybridSN is evaluated over three public available hyperspectral datasets on different amounts of training samples. Using only 5%, 1%, and 1% labeled data for training in Indian Pines, Salinas, and University of Pavia, respectively, the classification accuracy of R-HybridSN is 96.46%, 98.25%, 96.59%, respectively, which is far better than the contrast models.

An Improved 3D-2D Convolutional Neural Network Based on Feature Optimization for Hyperspectral Image Classification

Abstract: As a new technology in the field of remote sensing, hyperspectral remote sensing has been widely used in land classification, mineral exploration, environmental monitoring, and other areas. In recent years, deep learning has achieved outstanding results in hyperspectral image classification tasks. However, problems such as low classification accuracy for small sample classes in unbalanced datasets and lack of robustness of the models usually lead to unstable classification performance of hyperspectral images. Therefore, from the perspective of feature optimization, we propose an improved hybrid convolutional neural network for hyperspectral image feature extraction and classification. Different from the current simple multi-scale feature extraction, we first optimize the features of each scale, and then perform multi-scale feature fusion. To this end, we use 3D dilated convolution to design a multi-level feature extraction block (MFB), which can be used to extract features with different correlation strengths at a fixed scale. Then, we construct a spatial multi-scale interactive attention (SMIA) module in the spatial feature enhancement phase, which can refine the multi-scale features through the attention weights of multi-scale feature interaction, and further improve the quality of spatial features. Finally, experiments were performed on different datasets, including balanced and unbalanced samples. The results show that the proposed model is more accurate and the extracted features are more robust.

Recent Seasonal Spatiotemporal Variations in Alpine Glacier Surface Elevation in the Pamir

Abstract: Climate change can lead to seasonal surface elevation variations in alpine glaciers. This study first uses DEM (Digital Elevation Model) of Pamir glaciers to develop a denoising model for laser altimetry of ICESat-2 footprints, which reduces the standard deviation of the differences between ICESat-2 footprints and corresponding datum DEM from 13.9 to 3.6 m. Second, the study constructs a calibration processing model for solving the problem that laser footprints obtained at different times have inconsistent plane positions. We calculates plane position and elevation differences between the two laser footprints in the local area of 0.05 × 0.05° from 2018 to 2021. The elevations constructed by laser footprints shows a strong correlation with the datum elevation over the different periods, and effectively preserve the time-series variation information of glacier surface elevation (GSE). Based on these two models, the spatiotemporal variations of the surface elevation of the Pamir glaciers is established as a function of seasons. There are three main conclusions: (1) The GSE in the Pamir increased slightly from 2018 to 2021 at an average rate of +0.02 ± 0.01 m/year. The time series with elevation increase was located exactly on the glacial ablation zone, and the time series with elevation decrease occurred on the glacial accumulation zone. Both observations demonstrate the surge state of the glacier. (2) The Pamir eastern (Zone I) and northwestern (Zone III) regions had large glacier accumulation areas. GSE in these two regions has increased in recent years at yearly rates of +0.25 ± 0.13 and +0.06 ± 0.04 m/year, respectively. In contrast, the GSE of small glaciers in Zones II and IV has decreased at a yearly rate of −0.96 ± 0.37 and −0.24 ± 0.18 m/year, respectively. Climate was the primary factor influencing the increase in GSE in Zones I and III. The westerly circulation had been reinforced in recent years, and precipitation had increased dramatically at a rate of +0.99 mm/year in the northwestern section of the Pamir; this was the primary cause of the increase in GSE. (3) The increased precipitation and decreased temperature were both important factors causing an overall +0.02 ± 0.01 m/year variation of GSE in this region. The GSE in the four sub-regions showed different variation trends because of variations in temperature and precipitation. The external causes that affected the increase in GSE in the region included an average yearly temperature decrease at the rate of 0.54 ± 0.36 °C/year and a total yearly precipitation increase of 0.46 ± 0.29 mm/year in the study area from 2018 to 2021.

Benefits of depth in neural networks

Abstract: For any positive integer $k$, there exist neural networks with $\Theta(k^3)$ layers, $\Theta(1)$ nodes per layer, and $\Theta(1)$ distinct parameters which can not be approximated by networks with $\mathcal{O}(k)$ layers unless they are exponentially large --- they must possess $\Omega(2^k)$ nodes. This result is proved here for a class of nodes termed "semi-algebraic gates" which includes the common choices of ReLU, maximum, indicator, and piecewise polynomial functions, therefore establishing benefits of depth against not just standard networks with ReLU gates, but also convolutional networks with ReLU and maximization gates, sum-product networks, and boosted decision trees (in this last case with a stronger separation: $\Omega(2^{k^3})$ total tree nodes are required).

A New End-to-End Multi-Dimensional CNN Framework for Land Cover/Land Use Change Detection in Multi-Source Remote Sensing Datasets

Abstract: The diversity of change detection (CD) methods and the limitations in generalizing these techniques using different types of remote sensing datasets over various study areas have been a challenge for CD applications Additionally, most CD methods have been implemented in two intensive and time-consuming steps: (a) predicting change areas, and (b) decision on predicted areas In this study, a novel CD framework based on the convolutional neural network (CNN) is proposed to not only address the aforementioned problems but also to considerably improve the level of accuracy The proposed CNN-based CD network contains three parallel channels: the first and second channels, respectively, extract deep features on the original first- and second-time imagery and the third channel focuses on the extraction of change deep features based on differencing and staking deep features Additionally, each channel includes three types of convolution kernels: 1D-, 2D-, and 3D-dilated-convolution The effectiveness and reliability of the proposed CD method are evaluated using three different types of remote sensing benchmark datasets (ie, multispectral, hyperspectral, and Polarimetric Synthetic Aperture RADAR (PolSAR)) The results of the CD maps are also evaluated both visually and statistically by calculating nine different accuracy indices Moreover, the results of the CD using the proposed method are compared to those of several state-of-the-art CD algorithms All the results prove that the proposed method outperforms the other remote sensing CD techniques For instance, considering different scenarios, the Overall Accuracies (OAs) and Kappa Coefficients (KCs) of the proposed CD method are better than 9589% and 0805, respectively, and the Miss Detection (MD) and the False Alarm (FA) rates are lower than 12% and 3%, respectively

Spatial-Spectral Feature Refinement for Hyperspectral Image Classification Based on Attention-Dense 3D-2D-CNN.

Abstract: Convolutional neural networks provide an ideal solution for hyperspectral image (HSI) classification. However, the classification effect is not satisfactory when limited training samples are available. Focused on "small sample" hyperspectral classification, we proposed a novel 3D-2D-convolutional neural network (CNN) model named AD-HybridSN (Attention-Dense-HybridSN). In our proposed model, a dense block was used to reuse shallow features and aimed at better exploiting hierarchical spatial-spectral features. Subsequent depth separable convolutional layers were used to discriminate the spatial information. Further refinement of spatial-spectral features was realized by the channel attention method and spatial attention method, which were performed behind every 3D convolutional layer and every 2D convolutional layer, respectively. Experiment results indicate that our proposed model can learn more discriminative spatial-spectral features using very few training data. In Indian Pines, Salinas and the University of Pavia, AD-HybridSN obtain 97.02%, 99.59% and 98.32% overall accuracy using only 5%, 1% and 1% labeled data for training, respectively, which are far better than all the contrast models.

Multiscale Information Fusion for Hyperspectral Image Classification Based on Hybrid 2D-3D CNN

Abstract: Hyperspectral images are widely used for classification due to its rich spectral information along with spatial information. To process the high dimensionality and high nonlinearity of hyperspectral images, deep learning methods based on convolutional neural network (CNN) are widely used in hyperspectral classification applications. However, most CNN structures are stacked vertically in addition to using a onefold size of convolutional kernels or pooling layers, which cannot fully mine the multiscale information on the hyperspectral images. When such networks meet the practical challenge of a limited labeled hyperspectral image dataset—i.e., “small sample problem”—the classification accuracy and generalization ability would be limited. In this paper, to tackle the small sample problem, we apply the semantic segmentation function to the pixel-level hyperspectral classification due to their comparability. A lightweight, multiscale squeeze-and-excitation pyramid pooling network (MSPN) is proposed. It consists of a multiscale 3D CNN module, a squeezing and excitation module, and a pyramid pooling module with 2D CNN. Such a hybrid 2D-3D-CNN MSPN framework can learn and fuse deeper hierarchical spatial–spectral features with fewer training samples. The proposed MSPN was tested on three publicly available hyperspectral classification datasets: Indian Pine, Salinas, and Pavia University. Using 5%, 0.5%, and 0.5% training samples of the three datasets, the classification accuracies of the MSPN were 96.09%, 97%, and 96.56%, respectively. In addition, we also selected the latest dataset with higher spatial resolution, named WHU-Hi-LongKou, as the challenge object. Using only 0.1% of the training samples, we could achieve a 97.31% classification accuracy, which is far superior to the state-of-the-art hyperspectral classification methods.