A5059512471- Railway Engineering and Dynamics
- Topic Modeling
- Advanced Graph Neural Networks
- Civil and Geotechnical Engineering Research
- Semantic Web and Ontologies
- Natural Language Processing Techniques
- Geotechnical Engineering and Underground Structures
- Railway Systems and Energy Efficiency
- Data Quality and Management
- Autism Spectrum Disorder Research
- Complex Network Analysis Techniques
- Adhesion, Friction, and Surface Interactions
- Mechanical stress and fatigue analysis
- Biomedical Text Mining and Ontologies
- Domain Adaptation and Few-Shot Learning
- Text and Document Classification Technologies
- Sentiment Analysis and Opinion Mining
- Structural Health Monitoring Techniques
- Advanced Text Analysis Techniques
- Wikis in Education and Collaboration
- Vibration and Dynamic Analysis
- Recommender Systems and Techniques
- Graph Theory and Algorithms
- Service-Oriented Architecture and Web Services
- Artificial Intelligence in Law
Southeast University
2016-2025
Nanyang Technological University
2018-2024
Hangzhou Dianzi University
2023
China State Shipbuilding (China)
2023
Shanghai Jiao Tong University
2007-2019
University of Southampton
1999-2002
Misinformation has become a pressing issue. Fake media, in both visual and textual forms, is widespread on the web. While various deepfake detection text fake news methods have been proposed, they are only designed for single-modality forgery based binary classification, let alone analyzing reasoning subtle traces across different modalities. In this paper, we high-light new research problem multi-modal namely Detecting Grounding Multi-Modal Media Manipulation (DGM <sup...
SUMMARY For a high speed electrical rail system, good dynamic performance of the pantograph-catenary system is vital for smooth and continuous current collection. It has been known many years that to achieve this head pantograph should be made as light possible average stiffness catenary high. These conclusions, however, have reached by numerical modelling operational experience. In paper modelled time-varying, single degree-of-freedom facilitate an analytical investigation dynamics....
With the continuous development of intelligent technologies, knowledge graph, backbone artificial intelligence, has attracted much attention from both academic and industrial communities due to its powerful capability representation reasoning. In recent years, graph been widely applied in different kinds applications, such as semantic search, question answering, management so on. Techniques for building Chinese graphs are also developing rapidly have constructed support various applications....
Vision-language pretraining (VLP) has been investigated to generalize across diverse downstream tasks for fundus image analysis. Although recent methods showcase promising achievements, they significantly rely on large-scale private image-text data but pay less attention the manner, which limits their further advancements. In this work, we introduce MM-Retinal V2, a high-quality paired dataset comprising CFP, FFA, and OCT modalities. Then, propose novel vision-language model, namely KeepFIT...
Few-shot learning (FSL) has recently been extensively utilized to overcome the scarcity of training data in domain-specific visual recognition. In real-world scenarios, environmental factors such as complex backgrounds, varying lighting conditions, long-distance shooting, and moving targets often cause test images exhibit numerous incomplete or noise disruptions. However, current research on evaluation datasets methodologies largely ignored concept "environmental robustness", which refers...
Community search on attributed graphs (CSAG) is a fundamental topic in graph data mining. Given an G and query node q , CSAG seeks structural- attribute-cohesive subgraph from that contains . Exact methods based traversal are time-consuming, especially for large graphs. Approximate improve efficiency by pruning the space with heuristics but still take hundreds of milliseconds to tens seconds respond, hindering their use time-sensitive applications. Moreover, strategies typically tailored...
Abstract A study is presented of the non-linear dynamic interaction between a wheel and rail, excited by rough-ness on rail contact surfaces. moving irregularity model used to represent wheel/ process in time-domain. low order multiple degree-of-freedom system developed approximate infinite track for numerical simulations. The effects wheel/rail are investigated through calculations, analysis comparisons with results from linear model. difference interactions found be small if roughness...
Wheel/track impact due to rail joints or wheel flats has been studied over many years. The railway track is usually assumed be linear in order simplify the model, although pad and ballast are actually non-linear. This may cause incorrect results some extent circumstances, as stiffness varies with load. In this paper wheel/track using a non-linear model. represented by finite element (FE) model supported foundation. impacts simulated at different train speeds three types of pad, soft, medium...
The present work investigates corrugation growth and mitigation associated with vibration interference between multiple moving wheels resilient track. A discretely supported slab track is developed in the time domain. tangential contact wheel rail described using an unsteady two-dimensional model. Combined these two models long-term wear, processes of under single wheel-rail interaction are simulated. results show that low-frequency standing wave caused by vibratory interferences can lead to...
Recently, graph query is widely adopted for querying knowledge graphs. Given a G <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Q</sub> , the finds subgraphs in that exactly or approximately match . We face two challenges on query: (1) structural gap between and predefined schema causes mismatch with graph, (2) users cannot view answers until terminates, leading to longer system response time (SRT). In this paper, we propose semantic-guided...