A5053231956- Railway Engineering and Dynamics
- Infrastructure Maintenance and Monitoring
- Geotechnical Engineering and Underground Structures
- Civil and Geotechnical Engineering Research
- Structural Health Monitoring Techniques
- Railway Systems and Energy Efficiency
- Simulation and Modeling Applications
- Traffic Prediction and Management Techniques
- Mechanical stress and fatigue analysis
- Software Engineering Research
- Vehicle License Plate Recognition
- Handwritten Text Recognition Techniques
- Neural Networks and Reservoir Computing
- Technology and Security Systems
- Aquaculture disease management and microbiota
- Artificial Intelligence in Healthcare
- Advanced Sensor and Control Systems
- Software Reliability and Analysis Research
- Security and Verification in Computing
- Advanced Image and Video Retrieval Techniques
- Wireless Sensor Networks and IoT
- Vehicle Noise and Vibration Control
- Diverse Topics in Contemporary Research
- Transportation Planning and Optimization
- Time Series Analysis and Forecasting
China Academy of Railway Sciences
2020-2025
Northwest A&F University
2021-2024
China General Nuclear Power Corporation (China)
2021-2024
Energy Research Institute
2024
Huazhong University of Science and Technology
2021
Hubei Cancer Hospital
2021
Northeast Agricultural University
2015-2021
Zhejiang Lab
2021
Chengdu University of Technology
2021
Peking University
2020
Computer-assisted diagnosis is key for scaling up cervical cancer screening. However, current recognition algorithms perform poorly on whole slide image (WSI) analysis, fail to generalize diverse staining and imaging, show sub-optimal clinical-level verification. Here, we develop a progressive lesion cell method combining low- high-resolution WSIs recommend cells recurrent neural network-based WSI classification model evaluate the degree of WSIs. We train validate our analysis system 3,545...
Mature soybean phenotyping is an important process in breeding; however, the manual time-consuming and labor-intensive. Therefore, a novel approach that rapid, accurate highly precise required to obtain phenotypic data of stems, pods seeds. In this research, we propose mature phenotype measurement algorithm called Soybean Phenotype Measure-instance Segmentation (SPM-IS). SPM-IS based on feature pyramid network, Principal Component Analysis (PCA) instance segmentation. We also new method uses...
Micropterus salmoides rhabdovirus (MSRV) is an significant pathogen that causes high mortality and related economic losses in bass aquaculture. There no effective or approved therapy to date. In this study, we evaluated the anti-MSRV effects of 22 quinoline derivatives grass carp ovary (GCO) cells. Among these compounds, 8-hydroxyquinoline exhibited valid inhibition decreasing MSRV nucleoprotein gene expression levels 99.3% with a half-maximal inhibitory concentrations (IC50 ) value 4.66 μM...
Track supporting stiffness plays a crucial role in the running quality of vehicle crossing through turnout. This paper mainly focuses on influence random rubber tie plate vehicle-turnout coupling system. The relevant field is constructed via number theoretical method (NTM). coupled model with flexible rails built modal superposition method. probability density function (PDF) obtained evolution (PDEM). Cases normal distributions various mean values are compared to contrast impact dynamic...
The track regularity with high accuracy is crucial important to protect the dynamic safety in speed and heavy haul railways, normally, inertial measurement unit (IMU)is capable provide precise solution for irregularity measurement. However, achieve accuracy, most existing methods seriously depended on fiber-optic IMU, it very expensive, also larger size weight, while growing availability of low-cost small MEMS-IMU can yield acceptable by using multiple MEMS-IMUs fusion particular geometric...
Track geometry monitoring is essential for track maintenance. Dedicated inspection vehicles are scheduled to measure irregularities throughout the railway network, so cannot inspect each line frequently. It desirable much higher frequently using in-service trains. One possible way estimating from vehicle–body accelerations because accelerators can be easily installed in vehicle–body. However, inverting acceleration a pending issue. Up now, most research has focused on vehicle dynamics...
Scene text recognition has been a hot topic in computer vision. Recent methods adopt the attention mechanism for sequence prediction which achieve convincing results. However, we argue that existing faces problem of diffusion, model may not focus on certain character area. In this paper, propose Gaussian Constrained Attention Network to deal with problem. It is 2D attention-based method integrated novel Refinement Module, predicts an additional mask refine weights. Different from adopting...
Track irregularity detection serves as one of the most essential technologies to interpret in-service condition high-speed railway (HSR) and ensure comfortability safety in HSR service. This study newly proposed a novel on-board technique integrated with developed algorithm for identifying longitudinal track obtain monitoring further based maintenance (CBM). Such includes data acquisition unit, spatial-time synchronous calibration unit processing which can be directly installed commercial...
Ballast layer defects are the primary cause for rapid track geometry degradation. Detecting these in real-time during inspections is urgently needed to ensure safe train operations. To achieve this, an indicator, degradation rate (TDR) was proposed. This calculated using inspection data locate and predict railway-line sections with ballast defects. The TDR determined by monthly standard deviation of rail longitudinal level, which one aspect geometry. Layer Health Classification (BLHC)...
Though deep learning based scene text detection has achieved great progress, well-trained detectors suffer from severe performance degradation for different domains. In general, a tremendous amount of data is indispensable to train the detector in target domain. However, collection and annotation are expensive time-consuming. To address this problem, we propose self-training framework automatically mine hard examples with pseudo-labels unannotated videos or images. reduce noise examples,...
In this study, a wheel-rail transient rolling contact model capable of accounting for the nonlinear displacement-force properties hanging sleepers is proposed. The sleeper status affected by rail irregularities an input analysis behavior and related degradation in terms plastic deformation fatigue. results indicate that severity significantly geometric characteristics relative position with respect to surface irregularity. defects aggravate impact increase force amplitude, patch size,...
Abstract Vehicle‐mounted detection methods have been widely applied in the maintenance of high‐speed railways (HSRs), providing feasibility for diagnosing ballastless track arching. However, applying data faces several key limitations: (1) The threshold mostly requires manual setting, making recognition accuracy highly subjective; (2) extensive workload inspections makes it challenging to label data, hindering application supervised learning approaches. To address these problems, this paper...
Modern object detection methods based on convolutional neural network suffer from severe catastrophic forgetting in learning new classes without original data. Due to time consumption, storage burden and privacy of old data, it is inadvisable train the model scratch with both data when emerge after trained. In this paper, we propose a novel incremental detector Faster R-CNN continuously learn using It triple where an residual as assistants for helping previous learned knowledge. To better...
While the track irregularity in turnout areas has a significant impact on wheel-rail contact, driving safety, and stability, of front incoming vehicles is often overlooked. This paper fills gaps study. As result, rigid-flexible coupled dynamic model vehicle developed. The effect various irregularities performance at high speeds been investigated based random sampling method. results show that different types have effects responses vehicles. most sensitive to short-wavelength (3 m) turnout....
Accurate prediction of the change in track irregularity plays an essential role keeping high-speed railway safe and stable. Regular maintenance is important measure to guarantee smoothness, which has a great influence on state. This paper introduces time series anomaly detection model (AD) detect changes by tracking difference between mean values two sliding windows. These are taken as one holidays prophet called AD_ Prophet model. Moreover, this proposes multilevel residual (Re-Prophet)...