A5101821628- 3D Shape Modeling and Analysis
- Computer Graphics and Visualization Techniques
- Aortic Disease and Treatment Approaches
- Natural Language Processing Techniques
- Topic Modeling
- Aortic aneurysm repair treatments
- Generative Adversarial Networks and Image Synthesis
- Simulation and Modeling Applications
- Face recognition and analysis
- Video Surveillance and Tracking Methods
- Speech and Audio Processing
- Machine Learning and ELM
- Speech and dialogue systems
- Video Analysis and Summarization
- Security and Verification in Computing
- Music and Audio Processing
- Human Motion and Animation
- Face and Expression Recognition
- Teleoperation and Haptic Systems
- Network Security and Intrusion Detection
- Video Coding and Compression Technologies
- AI in cancer detection
- Advanced Computing and Algorithms
- Advanced Technologies in Various Fields
- Cardiac Structural Anomalies and Repair
University of Science and Technology Beijing
2016-2025
Xi’an Jiaotong-Liverpool University
2024
Qilu Hospital of Shandong University
2020-2023
East China Jiaotong University
2019-2023
Shandong University
2023
Dalian University
2012-2022
Beijing Electronic Science and Technology Institute
2022
Baidu (China)
2020
Chongqing University of Education
2020
Northwest Normal University
2020
Policy enforcement is an integral part of many applications. Policies are often used to control access sensitive information. Current policy specification languages give users fine-grained over when and how information can be accessed, flexible enough in a variety Evaluation these policies, however, not optimized for performance. Emerging applications, such as real-time privacy policies sensor network or location-aware computing environment, require high throughput. Our experiments indicate...
With the growing popularity of anomaly detection systems, which is due partly to rise in zero-day attacks, a new class threats have evolved where attacker mimics legitimate activity blend and avoid detection. We propose system called Siren that injects crafted human input alongside user thwart these mimicry attacks. The specially designed trigger known sequence network requests, compares actual traffic. It then flags unexpected messages as malicious. Using this method, we were able detect...
Due to the sustained increases in life expectancy over past half century, elderly today will receive supports from their children for a longer period than ever before. Therefore, understanding spillover effects of children's socioeconomic status on parents' health becomes increasingly important both scholars and policy makers.The Ordinary Least Squares regression is applied China Health Retirement Longitudinal Study 2011, national representative dataset including approximately 10 000...
Hyperhomocysteinemia (HHcy) is the risk factor for cardiovascular disease and stroke. However, impacts on genetic variation of methylene tetrahydrofolate reductase (MTHFR) plasma homocysteine levels in Northeast Chinese population have not been studied. Therefore, this study was carried out to determine relationship between HHcy MTHFR gene variation, whether it influenced by age sex China.A total 466 subjects were randomly enrolled study. According (Hcy ≥ 15 μmol/L) subjects, they divided...
Despite progress in the past decades, 3D shape acquisition techniques are still a threshold for various face-based applications and have therefore attracted extensive research. Moreover, advanced 2D data generation models based on deep networks may not be directly applicable to objects because of different dimensionality data. In this work, we propose two novel sampling methods represent faces as matrix-like structured that can better fit networks, namely (1) geometric method representation...
English feature recognition has a certain influence on the development of intelligent technology. In particular, speech technology problem accuracy when performing recognition. order to improve effect, this study takes learning algorithm as system and combines support vector machines construct an uses linear classifiers nonlinear complete relevant work subjective Moreover, spectral subtraction is introduced in front end extraction, amplitude noise-free signal subtracted from noise obtain...
Abstract As we are usually confronted with a large instance space for real-word data sets, it is significant to develop useful and efficient multiple-instance learning (MIL) algorithm. MIL, where training prepared in the form of labeled bags rather than instances, variant supervised learning. This paper presents novel MIL algorithm an extreme machine called MI-ELM. A radial basis kernel adapted approach problem using Hausdorff distance measure between bags. The clusters hidden layer composed...
With the recent advances in data acquisition techniques, compression of various dynamic mesh sequence has become an important topic computer graphics community. In this article, we present a new spatio-temporal segmentation-based approach for adaptive sequences. Given input sequence, first compute initial temporal cut to obtain small subsequence by detecting boundary behavior. Then, apply two-stage vertex clustering on resulting classify vertices into groups with optimal intra-affinities....
To evaluate the early remodeling of proximal aorta in patients with acute type B aortic dissection (ATBAD) after zone 2 thoracic endovascular repair (TEVAR).From January 2016 to May 2022, 53 ATBAD underwent TEVAR were divided into two groups, Castor single-branched stent-graft (CSS) group (n = 26) and common 27). Three-dimensional imaging created by computed tomography angiography was used measure different parameters aorta, such as angulation, cross-sectional area (CSA), length tortuosity....
With the recent advances of data acquisition techniques, compression various 3D mesh animation has become an important topic in computer graphics community. In this paper, we present a new spatio-temporal segmentation-based approach for animations. Given input sequence, first compute initial temporal cut to obtain small subsequence by detecting boundary dynamic behavior. Then, apply two-stage vertex clustering on resulting classify vertices into groups with optimal intra-affinities. After...
Although inflammation contributes to the development of acute kidney injury (AKI), role monocyte lymphocyte ratio (MLR) in predicting risk postoperative AKI patients underwent cardiac surgery is not completely clear yet. This study aimed investigate whether elevated MLR predicts AKI.In this retrospective study, a total 331 adult who heart valve replacement were included. These are distinguished between single or double replacement. was diagnosed, according KDIGO classification criteria. The...
We introduce a novel remote volume rendering pipeline for medical visualization targeted mHealth (mobile health) applications. The necessity of such stems from the large size imaging data produced by current CT and MRI scanners with respect to complexity volumetric algorithms. For example, resolution typical Angiography (CTA) easily reaches 512^3 voxels can exceed 6 gigabytes in spanning over time domain while capturing beating heart. This explosion makes transfers mobile devices...
Abstract T‐cell acute lymphoblastic leukemia (T‐ALL) is an aggressive malignant blood disorder with a high rate of relapse. Patients relapse as result minimal residual disease (MRD), which originates from T‐ALL cells in the bone marrow microenvironment (BMM). In present study, it observed that adipocytes increase dramatically BMM patients after exposure to chemotherapeutic drugs. Then, proved attract by releasing CXCL13 and support cell survival activating Notch1 signaling pathway via DLL1...
One of the most fundamental problems in computer security is protecting sensitive digital information from unauthorized disclosure. There are a number challenges, such as spyware, removable media, and mobile devices, which make this very hard problem. The problem becomes even more difficult when adversary somebody who authorized to view data. This what commonly referred an insider leak. Insider leaks often occur out malice, but sometimes just due plain negligence, was case with recent leak...
Large language models (LLMs) demonstrate remarkable multilingual capabilities without being pre-trained on specially curated parallel corpora. It remains a challenging problem to explain the underlying mechanisms by which LLMs process texts. In this paper, we delve into composition of Transformer architectures in pinpoint language-specific regions. Specially, propose novel detection method, activation probability entropy (LAPE), identify neurons within LLMs. Based LAPE, conduct comprehensive...
With urbanization reshaping the auditory landscape of cities, significance discerning hazardous sounds in outdoor environments has become paramount, especially for hearing-impaired community. This paper introduces an efficient Convolutional Neural Network (CNN) model tailored classification hazard to enhance safety individuals with hearing impairments. Our transforms audio signals into spectrogram images, leveraging power image techniques. Through rigorous training and optimization, achieves...
Diffusion Models (DMs) achieve state-of-the-art synthesis results in image generation and have been applied to various fields. However, DMs sometimes seriously violate user privacy during usage, making the protection of an urgent issue. Using traditional computing schemes like Secure Multi-Party Computation (MPC) directly faces significant computation communication challenges. To address these issues, we propose CipherDM, first novel, versatile universal framework applying MPC technology for...