A5003131468- 3D Shape Modeling and Analysis
- Image Processing and 3D Reconstruction
- Advanced Image and Video Retrieval Techniques
- Gastrointestinal disorders and treatments
- Advanced Image Processing Techniques
- Gastrointestinal motility and disorders
- Intestinal Malrotation and Obstruction Disorders
- Image Retrieval and Classification Techniques
- Advanced Vision and Imaging
- Esophageal and GI Pathology
- Simulation and Modeling Applications
- Image Processing Techniques and Applications
- Image and Signal Denoising Methods
- Image Enhancement Techniques
- Computer Graphics and Visualization Techniques
- Robotics and Sensor-Based Localization
- Enzyme Structure and Function
- Video Surveillance and Tracking Methods
- Natural Language Processing Techniques
- Topic Modeling
- Advanced Computational Techniques and Applications
- Advanced Neural Network Applications
- Geological Modeling and Analysis
- Computational Geometry and Mesh Generation
- Gastroesophageal reflux and treatments
Beijing Technology and Business University
2016-2025
Zhejiang University
2002-2024
Tsinghua University
2012-2024
Shandong Institute of Automation
2023
Chinese Academy of Sciences
2023
Beijing Academy of Artificial Intelligence
2023
University of Chinese Academy of Sciences
2023
Louisiana State University Health Sciences Center Shreveport
2021-2023
Shihezi University
2023
Xinjiang Production and Construction Corps
2023
Abstract Ulcerative colitis (UC), a chronic and nonspecific inflammatory disease of the intestine, has become prevalent global health concern. This guideline aims to equip clinicians caregivers with effective strategies for treatment management adult UC patients using traditional Chinese medicine (TCM). The systematically evaluated contemporary evidence through Grading Recommendations Assessment, Development, Evaluation framework. Additionally, it incorporated insights from ancient medical...
For heterogeneous federated learning, each client cannot ensure the reliability due to uncertainty in data collection, where different types of noise are always introduced into clients. Current existing methods rely on specific assumptions for distribution select clean samples or eliminate noisy samples. However, clients have deep neural network structures, and these models sensitivity various types, fixed noise-detection based may not be effective client. To overcome challenges, we propose...
Channel attention mechanism has been widely used in object detection algorithms because of its strong feature representation ability. The real-time algorithm YOLOv4 fast speed and high accuracy, but it still some shortcomings, such as inaccurate bounding box positioning poor robustness. Therefore, we introduced channel into the to enhance ability images, proposed a with mechanism. This module firstly carries out global average pooling operation on features extracted by YOLOv4, then local...
In order to achieve real-time updating of the domain knowledge graph and improve relationship extraction ability in construction process, a method is proposed. Based on structured Wikipedia’s classification system, we acquire concepts instances contained subject areas. A algorithm based co-word analysis intended extract relationships semi-structured open labels. Bi-GRU remote supervised model multiple-scale attention mechanism an improved cross-entropy loss function proposed obtain...
Although the use of computer color matching can reduce influence subjective factors by technicians, a natural tooth with ceramic restoration is still one most challenging topics in esthetic prosthodontics. Back propagation neural network (BPNN) has already been introduced into dentistry, but it disadvantages such as unstable and low accuracy. In our study, we adopt genetic algorithm (GA) to optimize initial weights threshold values BPNN for improving precision. To knowledge, firstly combine...
In modern medical diagnosis, the karyotype analysis for human chromosome is clinically significant diagnosis and treatment of genetic diseases. such an analysis, it critically important to segment banded chromosomes. Chromosome segmentation, however, technically challenging due variable features, complex background noise, uneven image quality images. Owing these technical challenges, existing deep-learning-based algorithms would have severe overfitting problems are ineffective in...
In order to solve the problems of low efficiency and subjectivity manual observation in process group-sheep-aggression detection, we propose a video streaming-based model for detecting aggressive behavior group sheep. experiment, collected videos sheep's daily routine sheep pen. Using open-source software LabelImg, labeled data with bounding boxes. Firstly, YOLOv5 detects all each frame outputs coordinates information. Secondly, sort using tracking heuristic proposed this paper. Finally,...
To efficiently compute arbitrary propellant grain evolution of the burning surface with uniform and non-uniform rate for solid rocket motor, a unified framework regression simulation has been developed based on minimum distance function. In order to speed up computation mini-mum between grid nodes triangular mesh surface, fast querying method equal size cube voxel structure was employed. An improved marching tetrahedron piecewise linear approximation carried out second-order tetrahedral...
Systemic corticosteroids represent the standard treatment for autoimmune pancreatitis with IgG4-associated cholangitis. For steroid-dependent disease, azathioprine has been used maintenance of remission. Mycophenolate mofetil transplant immunosuppression and more recently hepatitis; however, there are no case reports to date on use mycophenolate in adult patients pancreatitis. A patient IgG4-mediated cholangitis refractory steroids intolerant was treated mofetil, which inhibits de novo...
Background The respiratory syncytial virus (RSV) is the main cause of bronchiolitis in infants and interferon (IFN) α a commercial antiviral drug. nebulization IFN α1b could be viable treatment method. In this study, therapeutic effects safety delivery via infant were investigated multi-center prospective study. Methods findings Bronchiolitis patients admitted to 22 hospitals who met inclusion criteria enrolled randomly allocated four groups: control, Intramuscular Injection, Nebulization 1...
Most matting research resorts to advanced semantics achieve high-quality alpha mattes, and a direct low-level features combination is usually explored complement details. However, we argue that appearance-agnostic integration can only provide biased foreground (FG) details mattes require different-level feature aggregation for better pixel-wise opacity perception. In this article, propose an end-to-end hierarchical progressive attention network (HAttMatting++), which predict the of FG from...
With the development of computer vision and image processing, researchers have published abundant datasets for object detection. Whereas, we can hardly acquire food dataset dedicated To address this problem, introduce a novel that includes images 60 objects categories which are common in domain. total 78k labeled instances 60k images, creation our drew upon 30 graduate students lab involvement via multi-label automatic labeling system classification We present detailed statistical analysis...
Semantic similarity calculating has become a widespread topic in the field such as artificial intelligence, natural language processing, information retrieval and so on. WordNet is an English lexical database for synsets relations. It simple effective to calculate between concepts WordNet. Five semantic methods based on are discussed this paper. By analyzing comparing those methods, features of each presented experiment results given finally.