A5017408024- Mobile Learning in Education
- Online Learning and Analytics
- Data Mining Algorithms and Applications
- Educational Technology and Assessment
- Human Pose and Action Recognition
- Anomaly Detection Techniques and Applications
- Innovative Teaching and Learning Methods
- Network Security and Intrusion Detection
- Telemedicine and Telehealth Implementation
- Student Assessment and Feedback
- Infant Health and Development
- Reflective Practices in Education
- Power Systems and Renewable Energy
- Video Analysis and Summarization
- Artificial Immune Systems Applications
- Stroke Rehabilitation and Recovery
- Rough Sets and Fuzzy Logic
- Multimedia Communication and Technology
- Online and Blended Learning
- Advanced Authentication Protocols Security
- Acute Ischemic Stroke Management
- IoT and GPS-based Vehicle Safety Systems
- CO2 Reduction Techniques and Catalysts
- Higher Education Learning Practices
- Innovative Teaching Methods
University of Hong Kong
2015-2025
Chinese University of Hong Kong
2024
Open University of China
2015
Hong Kong University of Science and Technology
2015
Weakly supervised detection of anomalies in surveillance videos is a challenging task. Going beyond existing works that have deficient capabilities to localize long videos, we propose novel glance and focus network effectively integrate spatial-temporal information for accurate anomaly detection. In addition, empirically found approaches use feature magnitudes represent the degree typically ignore effects scene variations, hence result sub-optimal performance due inconsistency across scenes....
Classification and prediction of students' performance in examination are the typical challenges for educators. Various traditional data mining methods such as decision tree association rules were used to perform classification. In recent years, rapid development artificial intelligence deep learning algorithm provided another approach intelligent classification result prediction. this paper, a research on how use Tensorflow engine classifying forecasting their future universities degree...
Both educational data mining (EDM) and learning analytics (LA) focus on applying techniques to extract useful information from large sets. EDM is generally more interested in automated methods for discovery within the while LA relatively keen human-led understand involved processes. Among various fields of challenging studies EDM, domain structure aimed find knowledge an domain, such as formulating prerequisite requirements among components through online resources. However, with vast amount...
In the past, sports analysis is always conducted by specialized devices like sensors or depth camera, which are hard to promote. this essay, we use computer vision technology establish a sport detection and system with of an RGB camera deep learning models. It can recognize posture from image, make overall evaluation, return details real-time.The image will be processed two sub-systems sequentially: Human Performance system. The human used detect user's posture, while performance designed...
With the rapid development of distance education institution, there exist lots educational data to be analyzed for improving online teaching. Decision tree is an effective analysis algorithm in mining. In this paper, decision used analyze students' achievement provided by Open University China. The experiment result shows that students have different appearances between areas, and classify rules which can improve on-line teaching are obtained.
Radio frequency identification (RFID) technology is expected to become a critical and ubiquitous infrastructure of logistics supply chain management (SCM) related processes services. Low-cost has been the key RFID adoption in many logistics/SCM applications. However, deployment such tags may create new threats user privacy due powerful track trace capabilities tags, addition tag/reader manufacturing challenges. As result, some security mechanisms must be imposed on for addressing problems....
The paper proposes a cloud-based framework to abstract and analyze the meaningful rules among great amount of students' raw information. authors set learning skills based on course outline from Open University China. also present Apriori association algorithm rules, followed by reasonable analysis educational aspect supported k-means clustering identification.
The importance of Self-directed learning (SDL) in education is increasing the recent years.It helps promote students' lifelong attitudes and improve effectiveness.This stud evaluates efficiency effectiveness using SDL for teaching a group students with use an innovative mobile technology "iClass" which facilitates real-time interactive class via cloud-based platform.The results demonstrate that good solution to engage lesson uplift self-directed learning.
The development of Human Action Recognition (HAR) system is getting popular. This project developed a HAR for the application in surveillance to minimize man-power providing security citizens such as public safety and crime prevention. In this research, deep learning network using Recurrent Neural Network (RNN) with Long Short-Term Memory (LSTM) are used analyze dynamic video motion sport actions classify different types their performance. It could human small number frame efficiency memory...
Children's safety in the home has become an increasing social concern, and purpose of this experiment is to use machine learning detect child abuse scenarios order improve children's safety. This uses classify recognize voices predict whether current sound making crying, screaming or laughing. If a crying detected from child, immediate alert sent relevant personnel so that they can sense what may be experiencing respond timely manner blind spot surveillance. Together with hybrid video image...
In the past few years, many new mobile technologies including 3G, WiFi or mobileTV have created unprecedented learning opportunities on devices. Furthermore, such continuously fuel rapid growth of fields research like edutainment for educational entertainment. a recent project awarded by Hong Kong Wireless Development Center, we developed quiz game system 3G phone networks in China, other countries to facilitate anytime and anywhere. Our is so generic that it can be readily extended any...
Open education quality guarantee is a core issue in field of distance education. Data mining techniques are used to design effective teaching reform courses examination data analysis method would be good way for checking effects and could provide objective basis open assurance. This paper proposes solution based on Naive Bayesian classifier the impacts measures act quality. famous classifying method, as supervised learning, can extract valuable rules by using whose class label known train...
Quality management, radio frequency identification (RFID) and third generation mobile communications systems (3G) are increasing attracting the attention of corporate managers. This paper illustrates convergence three seemingly independent initiatives. RFID provides data inputs into company's quality improvement systems, such as Six Sigma, while 3G supports realtime between remote devices enterprise applications. readers can simultaneously scan identify hundreds tagged items. detailed give...
This research was a joint effort between the University of Hong Kong (HKU) and CLP, electricity power generation distribution service provider in Kong. The objective this to identify verify potential relationships performance standalone PV system, such as state charge battery output, other environmental factors solar radiation, wind speed, ambient temperature. results could help academia industry enhance knowledge on supply system under real operating conditions improve similar designs future.
The increasing energy supply is believed to exceed the capacity of three existing thermal power plants at Nansha Guangzhou for rapid development. Thus constructing an eco-friendly structure will be one main themes during region planning in near future. In this study, solar resources were assessed theoretically and simulated by PVsyst, PVGIS, Global Solar Atlas, RETScreen SSE Release 6.0 as well. It drew a conclusion that from 2010 2014, sun exposure hour percentage averaged 41.28%...
This study investigates the learning approach of designed Flipped Reading Platform (FRP) and its effects on primary school students' general Chinese reading comprehension capabilities. was undertaken as part Quality Education Fund project in Hong Kong, titled "Flipped Reading: Enhancing Learning Teaching Strategies Comprehension via an Interactive Cloud Platform."
Many lecturers in universities are held big lecture theater which may not be the most desirable environment to carry out effective class discussions. Class, an interactive mobile learning platform developed and adopted University of Hong Kong, enables real-time discussion among students lecturer through their tablet PCs smartphones facilitates classroom various sizes styles, gathering opinions from all classmates organized manner that is conducive fruitful It a useful tool facilitating...
Currently, the worldwide open and distance education institutions are carrying out teaching reform measures continuously to ensure quality, while courses examination data analysis results will be able reflect effectiveness objectively. So Wilcoxon Rank Sum Test based scheme is proposed in this paper detect change of courses. an effective Hypothesis testing method, which can deduce significant collectivity average value according sample by proposing hypothesis. It doesn't need know...
In recent years, the demand of in-class interaction and assessment for learning is rising. There more emphasis on using electronic tools in order to facilitate teachers seeking identify diagnose student problems, providing quality feedback students how improve their work. This paper discusses challenges e-assessment introduced a pen-based platform developed Hong Kong. The can provide an efficient effective channel so as monitor any difficulties help students' prior skills abilities, them...
The safety of children in home has become an increasing social concern, and the purpose this experiment is to use machine learning applied detect scenarios child abuse increase children. This uses classify recognize a child's voice predict whether current sound made by crying, screaming or laughing. If found be crying screaming, alert immediately sent relevant personnel so that they can perceive what may experiencing surveillance blind spot respond timely manner. Together with hybrid video...