A5057706808- Natural Language Processing Techniques
- Topic Modeling
- Particle accelerators and beam dynamics
- Gyrotron and Vacuum Electronics Research
- Web Data Mining and Analysis
- Time Series Analysis and Forecasting
- Music and Audio Processing
- Data Management and Algorithms
- Microwave Engineering and Waveguides
- Infrared Target Detection Methodologies
- Advanced SAR Imaging Techniques
- Anomaly Detection Techniques and Applications
- Transportation Planning and Optimization
- Algorithms and Data Compression
- Customer Service Quality and Loyalty
- Computational and Text Analysis Methods
- Traffic control and management
- Patient Satisfaction in Healthcare
- Quality and Supply Management
- Traffic Prediction and Management Techniques
- Optical Systems and Laser Technology
- Embedded Systems Design Techniques
- Pulsed Power Technology Applications
- Network Time Synchronization Technologies
- Complex Systems and Decision Making
University of Ottawa
2017-2024
AviChina Industry & Technology (China)
2021-2024
Huazhong University of Science and Technology
2006-2023
Wuhan National Laboratory for Optoelectronics
2023
Jiangxi Science and Technology Normal University
2010-2011
We report the design and initial experimental results of an 800-GHz second harmonic gyrotron prototype, in which a 15-T pulsed magnet is applied. This prototype developed at Wuhan National High Magnetic Field Center (WHMFC). An in-house software for analyzing designing terahertz gyrotrons research platform are developed. In addition, alignment adjustments, eddy current effects, radiation tests discussed. The show that can irradiate ~40 W fundamental 383.7-GHz (TE <sub...
In recent years, multimodal composite guidance technology has received extensive research and exploration, fusion methods based on deep learning have achieved performance far superior to traditional methods. However, in practical applications, there are still many difficulties the of infrared images high-resolution range profiles. response above challenges, this paper fully utilizes characteristics sensor data collection conduct in-depth application position prior, cross modal attention...
Traffic congestion has already been a distinctly serious problem in both developed and developing countries. Among the proposed methods to solve traffic problem, variable speed limit (VSL) is considered as one of most promising methods. But traditional VSL, sign control distance fixed, which makes VSL lack deployment flexibility. Whereas, connected vehicle (CV), crossing field intelligent transportation systems (ITS) internet things (IoT), flexibility can be achieved. Furthermore, big data...
It is predicted that fifty billion sensor-based devices are to be connected the Internet by 2020 with fast development of Things (IoT). Stream data mining on these tremendous has become an urgent task. Dynamic time warping (DTW) a popular similarity measure, which foundation stream mining. In last decade, DTW been well accelerated software and reconfigurable hardware optimizations. However, energy-efficiency not considered, critical for devices. this paper, we propose energy-efficient...
This article presents and evaluates a method for the detection of DBpedia types entities that can be used knowledge base completion maintenance. compares entity embeddings with traditional N-gram models coupled clustering classification. We tackle two challenges: (a) types, which to detect invalid assign type-less entities; (b) in resource description entity. Our results show outperform n-gram type contribute improvement DBpedia’s quality, maintenance, evolution.
Vehicle Re-identification is an effective way for calculation of travel time and origin-destination matrices, which critical traffic modeling optimization. Recently, dynamic warping (DTW) with magnetic signature has been a popular method vehicle re-identification. However, the increasing number vehicles real-time requirement estimation, DTW needs further acceleration. In this paper, we propose scalable parameterized architecture on reconfigurable fabrics efficient Parameterization adopted to...
The images of apple leaf disease under natural conditions has a complex background and small diseased spots, which makes it difficult to detect in real time. To address this, new method was proposed for spots segmentation based on DeepLabV3+ network. We used Xception composed depthwise separable convolutional layers as feature extraction Depthwise can effectively reduce the amount model parameters improve training efficiency. introduction CBAM module allows better focus spot area, influence...
In the gyrotron, phenomenon that electron velocity spread coupled with beam radius may occur, which is quite different from traditionally considered uncoupled ones. this paper, formation of such spatial investigated. It found inhomogeneous electric and magnetic fields in cathode region are responsible for problem. This much more visible low-voltage, high-frequency gyrotrons, Larmor be significantly smaller than spread.
<title>Abstract</title> The examination of gender bias, alongside other demographic biases like race, nationality, and religion, within generative large language models (LLMs), is increasingly capturing the attention both scientific community industry stakeholders. These often permeate LLMs, influencing widely used products potentially compromising user experiences. A growing body research dedicated to enhancing representations in natural processing (NLP) across a spectrum LLMs. This paper...
The examination of gender bias, alongside other demographic biases like race, nationality, and religion, within generative large language models (LLMs), is increasingly capturing the attention both scientific community industry stakeholders. These often affect LLMs, influencing popular products potentially compromising user experiences. A growing body research dedicated to enhancing representations in natural processing (NLP) across a spectrum LLMs. This paper explores current focused on...
In the paper, authors proposed concept of patient satisfaction degree and established some factors that affect according to research on spot in concrete hospitals systemic analysis. Firstly, it took ten 14 three A-level as examples random, handed out questionnaires degree. Then paper easily obtained initial information table based collation analysis collected questionnaires. At last, classified these tables by means k-means cluster through software SPSS 16.0. From result analysis, were...
This paper combines the advantages of regular segmentation and statistical segmentation, proposing an incomplete maximum matching method based on semantics. On basis ensuring time consumption, new solves defect word adhesion in algorithm. The innovative work includes: preprocessing stage, forward semantic similarity dictionary is constructed to realize follow-up recognition. In stage formula three feature weight proposed redefine principle. experimental results show that has a certain...
The paper takes system dynamics method as study tool, based on the research in specific hospital and analysis about competitiveness, combined with actual situation of Jiujiang Chinese hospital, built up seven simulation models, according to history data future development plans, established each variable equation simulated through software predicted trends which can provide a reliable basis for decision-making administrators related administrative departments.
In this paper, a Quasi-Optical launcher operating at 263 GHz with low order TE3,2 mode is designed for THz gyrotron Wuhan National High Magnetic Center (WHMFC). A to TE <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">8,3</inf> converter up front, because the lower-order does not follow selection rules Denisov-type launchers. The consists of left-handed (LH) right-handed (RH) xmlns:xlink="http://www.w3.org/1999/xlink">8,2</inf> 11-fold helically...