A5003378750- Advanced Neural Network Applications
- Adversarial Robustness in Machine Learning
- Remote Sensing and Land Use
- Underwater Acoustics Research
- Smart Agriculture and AI
- Medical Image Segmentation Techniques
- Underwater Vehicles and Communication Systems
- Advanced SAR Imaging Techniques
- Advanced Image and Video Retrieval Techniques
- Advanced Computational Techniques and Applications
- Visual Attention and Saliency Detection
- Anomaly Detection Techniques and Applications
- Advanced Image Processing Techniques
- Image Enhancement Techniques
- Remote Sensing and LiDAR Applications
- Remote-Sensing Image Classification
- Network Security and Intrusion Detection
- Domain Adaptation and Few-Shot Learning
- Hydrocarbon exploration and reservoir analysis
- Radar Systems and Signal Processing
- Radiomics and Machine Learning in Medical Imaging
- Remote Sensing in Agriculture
- Topic Modeling
- Advanced Algorithms and Applications
- Advanced Malware Detection Techniques
Xijing University
2015-2025
Northwestern Polytechnical University
2010-2025
Dalian Medical University
2021-2025
Qiqihar Medical University
2025
Shanghai Electric (China)
2024
Chinese Academy of Sciences
2001-2024
Sinopec (China)
2017-2024
Guilin University of Technology
2024
Hangzhou Dianzi University
2021-2024
Beijing Institute of Technology
2022-2024
The introduction of deep learning (DL) technology can improve the performance cyber–physical systems (CPSs) in many ways. However, this also brings new security issues. To tackle these challenges, article explores vulnerabilities DL-based unmanned aerial vehicles (UAVs), which are typical CPSs. Although research works have been reported previously on adversarial attacks DL models, only few them concerned about safety-critical CPSs, especially regression models such systems. In article, we...
Aircraft detection in synthetic aperture radar (SAR) images plays an essential role satellite observation and military decisions. Due to discrete scattering properties, speckle noise interference, various aircraft types, many existing methods struggle achieve the desired performance. In this article, we propose innovative semantic condition constraint guided feature aware network (SCFNet) for detecting different categories SAR images. First, considering properties of aircraft, design a...
Measurement-device-independent quantum key distribution (MDI-QKD) can eliminate all detector side channels and it is practical with current technology. Previous implementations of MDI-QKD used two symmetric similar losses. However, the secret rate severely limited when different have Here we report results first high-rate experiment over asymmetric channels. By using recent 7-intensity optimization approach, demonstrate>10×higher than previous best-known protocols for in situation large...
Point cloud classification plays a critical role in point processing and analysis. Accurately classifying objects on the ground urban environments from airborne laser scanning (ALS) clouds is challenge because of their large variety, complex geometries, visual appearances. In this paper, novel framework presented for effectively extracting shape features an ALS cloud, then, it used to classify small cloud. framework, split into hierarchical clusters different sizes based natural exponential...
Although state estimation using a bad data detector (BDD) is key procedure employed in power systems, the vulnerable to false injection attacks (FDIAs). Substantial deep learning methods have been proposed detect such attacks. However, neural networks are susceptible adversarial or examples, where slight changes inputs may lead sharp corresponding outputs even well-trained networks. This article introduces joint example and FDIAs (AFDIAs) explore various attack scenarios for systems....
The efficiency of some dimensionality reduction techniques, like lung segmentation, bone shadow exclusion, and t-distributed stochastic neighbor embedding (t-SNE) for exclusion outliers, is estimated analysis chest X-ray (CXR) 2D images by deep learning approach to help radiologists identify marks cancer in CXR. Training validation the convolutional neural network (CNN) was performed on open JSRT dataset (dataset #01), after - BSE-JSRT #02), segmentation #03), #04), segmented outliers t-SNE...
Vulnerability of various machine learning methods to adversarial examples has been recently explored in the literature. Power systems which use these vulnerable face a huge threat against examples. To this end, we first propose signal-specific method and universal signal-agnostic attack power using generated Second, black-box attacks based on transferable characteristics above two are also proposed evaluated. Third, training is adopted defend attacks. Experimental analyses demonstrate that...
State estimation methods used in cyber–physical systems (CPSs), such as smart grid, are vulnerable to false data injection attacks (FDIAs). Although substantial deep learning have been proposed detect attacks, neural networks (DNNs) highly susceptible adversarial which modify input of DNNs with unnoticeable but malicious perturbations. This article proposes a method explore targeted and stealthy FDIAs via machine learning. We pose sparse optimization problems achieve initial attack...
Automatic underwater target detection plays a vital role in sonar image processing and analysis, its core task is to discriminate categories achieve precise positioning. However, the interfered by seafloor reverberation noise complex background, which brings more significant challenges accurate of target. To different targets image, we proposed an adaptive global feature enhancement network (AGFE-Net), uses multi-scale convolution attention mechanisms with receptive field obtain semantic...
Side-scan sonar is an important application in the field of ocean exploration. Accurate segmentation target regions side-scan images a challenging issue due to low-resolution and strong noise interference. To accurately faster segment different categories image, novel convolutional neural networks (CNNs) model proposed this study. Firstly, deep separable residual module used for multi-scale feature extraction suppression information interference, multi-channel fusion method enhance transfer...
Despite recent works that have achieved remarkable progress on salient object detection for natural scene images, to detect various types and scales of objects, complex backgrounds in remote sensing images are still challenging. In this study, a novel global perception network (GPNet) is constructed the images. The proposed GPNet includes module (GPM), an axial attention block (AAB), feature distillation structure (FDS). GPM used preserve relationships entire dataset, AAB designed capture...
Sonar image object detection is essential in underwater rescue and resource exploration. Although many convolution neural network (CNN)-based algorithms have achieved great success natural images. However, for sonar images, problems, such as seabed reverberation noise interference, low proportion of foreground region pixels, poor imaging resolution, present considerable challenges to achieving accurate detection. To address these we propose a novel detector called the multilevel feature...
Abstract Deep learning models are easily deceived by adversarial examples, and transferable attacks crucial because of the inaccessibility model information. Existing SOTA attack approaches tend to destroy important features objects generate examples. This paper proposes split grid mask (SGMA), which reduces intensity model-specific transformation, effectively highlighting input image. Perturbing these can guide development examples in a more direction. Specifically, we introduce...
For semantic segmentation of remote sensing images, convolutional neural networks (CNNs) have proven to be powerful tools. However, the existing CNN-based methods problems feature information loss, serious interference by clutter information, and ignoring correlation between different scale features. To solve these problems, this article proposes a novel hidden feature-guided network (HFGNet) for which achieves accurate hierarchically extracting fusing valuable information. Specifically,...
Liver and tumor segmentation is an important technology for the diagnosis of hepatocellular carcinoma. However, most existing methods struggle to accurately delineate boundaries liver due significant differences in their shapes, sizes, distributions, which leads unclear contour incorrect delineation lesion area. To address this gap, we propose a hybrid gabor attention convolution transformer interaction network with hierarchical monitoring mechanism segmentation, named HyborNet. Generally,...
Predicting protein-ligand binding affinities is a critical problem in drug discovery and design. A majority of existing methods fail to accurately characterize exploit the geometrically invariant structures complexes for predicting affinities. In this study, we propose Geo-protein-ligand affinity (PLA), geometric equivariant graph representation learning framework with local-global structure awareness, predict by capturing information complexes. Specifically, local structural 3-D extracted...
Chronic hepatitis B (CHB) is a common cause of liver cirrhosis (LC), condition associated with an unfavourable prognosis. Therefore, timely diagnosis LC in CHB patients crucial. This study aimed to enhance the diagnostic accuracy by integrating stiffness measurement (LSM) traditional indicators. The participants were randomly divided into training and internal validation sets. Employing least absolute shrinkage selection operator (LASSO) random forest-recursive feature elimination (RF-RFE)...
Small non-coding molecules known as microRNAs (miRNAs) play a critical role in disease diagnosis, treatment, and prognosis evaluation. Traditional wet-lab methods for validating miRNA–disease associations are often time-consuming inefficient. With the advancement of high-throughput sequencing technologies, deep learning have become effective tools uncovering potential patterns revealing novel biological insights. Most existing approaches focus primarily on individual molecular behavior,...