A5016645765- Adversarial Robustness in Machine Learning
- Video Surveillance and Tracking Methods
- Anomaly Detection Techniques and Applications
- Functional Brain Connectivity Studies
- Face recognition and analysis
- Advanced Image and Video Retrieval Techniques
- Advanced Neural Network Applications
- Generative Adversarial Networks and Image Synthesis
- Face and Expression Recognition
- Digital Media Forensic Detection
- Neural dynamics and brain function
- Medical Image Segmentation Techniques
- Network Security and Intrusion Detection
- Advanced Malware Detection Techniques
- AI in cancer detection
- Multimodal Machine Learning Applications
- Image and Video Stabilization
- EEG and Brain-Computer Interfaces
- Domain Adaptation and Few-Shot Learning
- Advanced Vision and Imaging
- Image Retrieval and Classification Techniques
- Remote-Sensing Image Classification
- Advanced Measurement and Detection Methods
- Advanced Image Processing Techniques
- Infrared Target Detection Methodologies
Nanjing University of Science and Technology
2016-2025
Zhejiang Normal University
2025
Xuzhou Central Hospital
2024
University of Science and Technology Beijing
2022
Xi'an Technological University
2017-2019
Yale University
2013-2015
University of Connecticut
2013
Nanyang Technological University
2010-2012
Shenyang Center for Disease Control and Prevention
2011
Jilin University
2007-2008
Background. Glucocorticoids increase the risk of developing critical disease from viral infections. However, primary care practitioners in China use them as antipyretics, potentially exposing hundreds millions to this risk. Methods. We enrolled all patients with confirmed pandemic influenza A (pH1N1) virus infection aged ≥3 years available medical records at 4 Shenyang City hospitals 20 October 30 November 2009. patient was any confirmed, hospitalized pH1N1 who developed ≥1 following: death,...
In this paper, we propose a novel contour-seed pairs learning-based framework for robust and automated cell/nucleus segmentation. Automated granular object segmentation in microscopy images has significant clinical importance pathology grading of the cell carcinoma gene expression. The focus past literature is dominated by either segmenting certain type cells/nuclei or simply splitting clustered objects without contours inference them. Our method addresses these issues formulating detection...
In the era of rapid development information technology, it is particularly important to ensure security systems. The network attack surface, as an index for measuring system security, has become focus practitioners. At present, accuracy and practicability surface evaluations are insufficient. order solve this problem, paper proposes a evaluation method based on optimal strategy. This first identifies main targets resources then uses advanced optimization techniques determine best Finally,...
The Walrus Optimization (WO) algorithm, as an emerging metaheuristic has shown excellent performance in problem-solving, however it still faces issues such slow convergence and susceptibility to getting trapped local optima. To this end, the study proposes a novel WO enhanced by quasi-oppositional-based learning chaotic search mechanisms, called QOCWO. aims prevent premature optima enhance diversity of population integrating mechanism into original thereby improving global capability...
Abstract Background Transcription factors (TFs) regulate the genes’ expression by binding to DNA sequences. Aligned TFBSs of same TF are seen as cis-regulatory motifs, and substantial computational efforts have been invested find motifs. In recent years, convolutional neural networks (CNNs) succeeded in TF-DNA prediction, but existing DL methods’ accuracy needs be improved convolution function prediction should further explored. Results We develop a cascaded network model named CacPred...
Multivariate connectivity and functional dynamics have been of wide interest in the neuroimaging field, a variety methods developed to study interactions dynamics. In contrast, temporal dynamic transitions multivariate among brain networks, particular, resting state, much less explored. This article presents novel Bayesian variable partition model (DBVPM) that simultaneously considers models their via unified framework. The basic idea is detect boundaries piecewise quasi-stable interaction...
Abstract Modeling abnormal temporal dynamics of functional interactions in psychiatric disorders has been great interest the neuroimaging field, and thus a variety methods have proposed so far. However, disease‐related abnormalities within specific data‐driven discovered subnetworks rarely explored yet. In this work, we propose novel computational framework composed an effective Bayesian connectivity change point model for modeling brain their simultaneously variant nonnegative matrix...
Recently, correlation filters (CFs) for visual tracking present competitive performances on both accuracy and robustness, but there is still a need improving their overall capabilities. Most CF trackers learn best filter to regress training data fixed target response, which might lead drifting. In this paper, we an appealing tracker based the Kernelized Correlation Filter (KCF), can adaptively change response. Furthermore, utilize fast accurate scale estimation approach by learning...
Mobile crowdsourcing is a new computing paradigm that enables outsourcing computation tasks to mobile crowd nodes by means of offloading the from user edge (MEC) server. This article studies problem scheduling security-critical applications in multiserver MEC environment. We formulate this as an integer program and propose family convergent grey wolf optimizer (CGWO) metaheuristic algorithms seek for best solutions. Our proposed CGWO uses task permutation represent candidate solution...
Saliency maps have proven to be a highly efficacious approach for explicating the decisions of convolutional neural networks (CNNs). However, extant methodologies predominantly rely on gradients, which constrain their ability explicate complex models. Furthermore, such approaches are not fully adept at leveraging negative gradient information improve interpretive veracity. In this study, we present novel concept, termed positive and excitation (PANE), enables direct extraction PANE each...
In real-world surveillance systems, the person images captured by camera network consists of various low-resolution (LR) images. It creates a resolution mismatching problem when compared against high-resolution targeted person. significantly affects performance re-Identification. This is known as Low-Resolution Person re-identification (LR PREID). An efficient strategy would be to exploit image super-resolution (SR) with mutual learning approach. this paper, we propose novel method...
This paper presents a novel method for automated morphology delineation and analysis of cell nuclei in histopathology images. Combining the initial segmentation information concavity measurement, proposed first segments clusters into individual pieces, avoiding errors introduced by scale-constrained Laplacian-of-Gaussian filtering. After that nuclear boundary-to-marker evidence computing is to delineate objects after refined process. The obtained set then modeled periodic B-splines with...
Multiple recent neuroimaging studies have demonstrated that the human brain's function undergoes remarkable temporal dynamics. However, quantitative characterization and modeling of such functional dynamics been rarely explored. To fill this gap, we presents a novel Bayesian connectivity change point model (BCCPM), to analyze joint probabilities among nodes brain networks between different time periods statistically determine boundaries blocks estimate points. Intuitively, determined points...
In the explainable artificial intelligence (XAI) field, an algorithm or a tool can help people understand how model makes decision. And this to select important features reduce computational costs realize high-performance computing. But existing methods are usually used visualize highlight active neurons, and few of them show importance relationships between features. recent years, some based on white-box approach have taken into account, but most only work specific models. Although...
Abstract Background The Assay for Transposase-Accessible Chromatin using sequencing (ATAC-seq) utilizes the Transposase Tn5 to probe open chromatic, which simultaneously reveals multiple transcription factor binding sites (TFBSs) compared traditional technologies. Deep learning (DL) technology, including convolutional neural networks (CNNs), has successfully found motifs from ATAC-seq data. Due limitation of width kernels, existing models only find with fixed lengths. A Graph network (GNN)...