A5100749027- Text and Document Classification Technologies
- Social Robot Interaction and HRI
- Assistive Technology in Communication and Mobility
- Extracellular vesicles in disease
- Advanced Graph Neural Networks
- Robotic Locomotion and Control
- Cancer-related molecular mechanisms research
- Advanced Text Analysis Techniques
- Reinforcement Learning in Robotics
- Rough Sets and Fuzzy Logic
- Complex Network Analysis Techniques
- Robot Manipulation and Learning
- AI in Service Interactions
- Handwritten Text Recognition Techniques
- Tumors and Oncological Cases
- Integrated Energy Systems Optimization
- Advanced Computational Techniques and Applications
- Advanced Computing and Algorithms
- Control and Dynamics of Mobile Robots
- Ethics in Clinical Research
- Advanced Malware Detection Techniques
- Electric Power System Optimization
- Sports and Physical Education Research
- Mesenchymal stem cell research
- Data Stream Mining Techniques
Imperial College London
2024
Southwest Jiaotong University
2024
China People's Public Security University
2024
Huazhong University of Science and Technology
2023
Affiliated Hospital of Nantong University
2022-2023
Nantong University
2022-2023
Georgia Institute of Technology
2020
Shanghai Children's Hospital
2015
Shanghai Jiao Tong University
2015
Anyang Normal University
2007
Graph Neural Networks (GNNs) play a significant role in tasks related to homophilic graphs.Traditional GNNs, based on the assumption of homophily, employ low-pass filters for neighboring nodes achieve information aggregation and embedding.However, heterophilic graphs, from different categories often establish connections, while same category are located further apart graph topology.This characteristic poses challenges traditional leading issues "distant node modeling deficiency" "failure...
Robot-assisted bed bathing holds the potential to enhance quality of life for older adults and individuals with mobility impairments. Yet, accurately sensing human body in a contact-rich manipulation task remains challenging. To address this challenge, we propose multimodal approach that perceives 3D contour parts using visual modality while capturing local contact details tactile modality. We employ Transformer-based imitation learning model utilize information learn focus on crucial...
It is particularly necessary for biomedical researchers to obtain applicable biosamples accurately and efficiently, especially from a biobank with multiple-disease catalogs. To optimize the retrieval procedure, in early stages of non-automatic biobank, we developed procedure that combined electronic information system graphically designed printed recording system, which assisted retrieving samples quickly visualized way. In this tables depending on structure equipment registered...
Versatile robotic caregivers could benefit millions of people worldwide, including older adults and with disabilities. Recent work has explored how can learn to interact through physics simulations, yet transferring what been learned real robots remains challenging. Virtual reality (VR) the potential help bridge gap between simulations world. We present Assistive VR Gym (AVR Gym), which enables virtual assistive robots. also provide evidence that AVR researchers improve performance...
The temporomandibular joint (TMJ) is a complex consisting of the condyle, temporal articular surface, and disc. Functions such as mastication, swallowing articulation are accomplished by movements TMJ. To date, TMJ has been studied more extensively, but types cells, their differentiation, interrelationship during growth development still unclear study limited. aim this was to establish molecular cellular atlas human embryonic condyle (TMJC) single-cell RNA sequencing, which will contribute...
Traditional transmission network expansion planning models are mostly based on the DC model. Compared with AC model, model is simple and easy to deal with, but it has many disadvantages. With extensive use of new energy, grid-connected capacity energy sources getting higher, which introduce uncertainty power grid. Therefore, necessary study wind photovoltaic energy. This paper proposes a multi-objective considers safety PV flow. The first objective function investment cost line expansion,...
In recent years, Unmanned Surface Vehicles (USVs) have been widely used in ocean transportation, search and rescue operations other fields. This article focuses on the formation navigation problem of multi-USV clusters. We deep reinforcement learning methods to establish a model for designed reward function which guides each USV its expected position while flexibly avoiding obstacles. By introducing virtual leader, cluster successfully follow path maintaining The simulation results...
Versatile robotic caregivers could benefit millions of people worldwide, including older adults and with disabilities. Recent work has explored how can learn to interact through physics simulations, yet transferring what been learned real robots remains challenging. Virtual reality (VR) the potential help bridge gap between simulations world. We present Assistive VR Gym (AVR Gym), which enables virtual assistive robots. also provide evidence that AVR researchers improve performance...
Currently, unmanned underwater vehicle (UUV) clusters have demonstrated effectiveness in performing oceanic tasks, with heterogeneity of UUVs becoming a prevalent trend development. Therefore, selecting appropriate heterogeneous members is essential for achieving efficient and high-quality task execution. To address this issue, study proposes UUV cluster cooperative system assignment method based on graph neural networks (GNN). This models the as graphical structure utilizes GNN-based...
This paper is to explore the combining theory and practice on training outstanding doctors, further investigate student-centered, ability-oriented innovation ability teaching philosophy. During course of teaching, problem-based learning (PBL) combined with case-based (CBL) experimental epidemiological scientific research were applied. All these enable clinical medical students master basic concepts, principle method epidemiology. The reformed curriculum made: excellent rate PBL reach76.8%...
Humans can leverage hierarchical structures to split a task into sub-tasks and solve problems efficiently. Both imitation reinforcement learning or combination of them with have been proven be an efficient way for robots learn complex tasks sparse rewards. However, in the previous work learning, tested environments are relatively simple 2D games, action spaces discrete. Furthermore, many works focusing on improving policies learned from expert polices that hard-coded trained by algorithms,...
Abstract Background: The temporomandibular joint (TMJ) is a complex consisting of the mandibular condyle, temporal articular surface, and disc. functions mastication, swallowing articulation are accomplished by movements TMJ. To date, TMJ has been studied more extensively, but study limited type cells, their differentiation, interrelationship during growth development unclear. aim this to establish molecular cellular developmental atlas human single-cell RNA sequencing, which will contribute...