A5102816667- Software Engineering Research
- Software Reliability and Analysis Research
- Metaheuristic Optimization Algorithms Research
- Software System Performance and Reliability
- Advanced Multi-Objective Optimization Algorithms
- Hand Gesture Recognition Systems
- E-commerce and Technology Innovations
- Non-Destructive Testing Techniques
- Industrial Vision Systems and Defect Detection
- High-Voltage Power Transmission Systems
- HVDC Systems and Fault Protection
- Software Engineering Techniques and Practices
- Adsorption and biosorption for pollutant removal
- Advanced Database Systems and Queries
- Domain Adaptation and Few-Shot Learning
- Human Pose and Action Recognition
- Advanced Sensor and Control Systems
- Vibrio bacteria research studies
- Railway Systems and Energy Efficiency
- Advanced Neural Network Applications
- Power System Reliability and Maintenance
- Berberine and alkaloids research
- Transport Systems and Technology
- Humor Studies and Applications
- Advanced Clustering Algorithms Research
Wuhan University of Science and Technology
2020-2025
Guangdong University of Foreign Studies
2019-2024
City University of Macau
2024
Hunan University of Traditional Chinese Medicine
2023
Shanghai University of Engineering Science
2017-2021
Southwest Jiaotong University
2020
Dalian Neusoft University of Information
2019
South China University of Technology
2018-2019
Shanghai University
2017
University of Washington
2013
This demo presents WiSee, a novel human-computer interaction system that leverages wireless networks (e.g., Wi-Fi), to enable sensing and recognition of human gestures motion. Since wire- less signals do not require line-of-sight can traverse through walls, WiSee enables interfaces for remote device control building automation. Further, it achieves this goal without requiring instrumentation the body with devices. We integrate applications demonstrate how users use including music players...
Cross-project defect prediction (CPDP) is a practical solution that allows software (SDP) to be used earlier in the lifecycle. With CPDP technique, predictor trained by labeled data of mature projects can applied for task new project. Most previous approaches ignored semantic information source code, and existing semantic-feature-based SDP methods do not take into account distribution divergence between projects. These limitations may weaken performance. To solve these problems, we propose...
Machine-learning-based software defect prediction (SDP) methods are receiving great attention from the researchers of intelligent engineering. Most existing SDP performed under a within-project setting. However, there usually is little to no training data learn an available supervised model for new task. Therefore, cross-project (CPDP), which uses labeled source projects predictor target project, was proposed as practical solution. In real CPDP tasks, class imbalance problem ubiquitous and...
Molybdenum disulfide (MoS2) has been widely applied in photocatalysts, field-effect transistors (FETs), and solar cells virtue of its high specific surface area superior carrier mobility. Nevertheless, conventional MoS2-based FETs involve Au Ag as the electrode materials, resulting a manufacturing cost. More importantly, preparation methods (e.g., chemical vapor deposition) require high-temperature process transfer MoS2, well subsequent deposition, for device fabrication, while MoS2 tends to...
Expensive optimization problems arise in diverse fields, and the expensive computation terms of function evaluation poses a serious challenge to global algorithms. In this article, simple yet effective algorithm for computationally is proposed, which called neighborhood regression algorithm. For minimization problem, proposed incorporates technique based on structure predict descent direction. The direction then adopted generate new potential offspring around best solution obtained so far....
Software defect prediction (SDP) technology is receiving widely attention and most of SDP models are trained on data from the same project. However, at an early phase software lifecycle, there little to no within-project training learn available supervised defect-prediction model. Thus, cross-project (CPDP), which learning a predictor for target project by using labelled source project, has shown promising value in SDP. To better perform CPDP, current studies focus filtering instances or...
Although the machine learning-based software defect prediction (SDP) method has shown promising value in engineering, yet challenges remain.To improve performance of SDP, some researchers have used deep learning algorithms to extract semantic and structural features program.However, more practical cross-project (CPDP) tasks, whether learning-generated can be directly should explored due data distribution shift that usually exists different projects.In this paper, we propose a Transferable...
Using classification methods to predict software defect is receiving a great deal of attention and most the existing studies primarily conduct prediction under within-project setting. However, there usually had no or very limited labelled data train an effective model at early phase lifecycle. Thus, cross-project (CPDP) proposed as alternative solution, which learning predictor for target project by using from source project. Differing previous CPDP that mainly apply instances selection...
The rapid rise of real-time bidding-based online advertising has brought significant economic benefits and attracted extensive research attention. From the perspective an advertiser, it is crucial to perform accurate utility estimation cost for each individual auction in order achieve cost-effective advertising. These problems are known as click through rate (CTR) prediction task market price modeling task, respectively. However, existing approaches treat CTR two independent tasks be...
Cross-project defect prediction (CPDP) is an important research direction in software prediction. Traditional CPDP methods based on hand-crafted features ignore the semantic information source code. Existing deep learning model may not fully consider differences among projects. Additionally, these accurately classify samples near classification boundary. To solve problems, authors propose a multi-adaptation and nuclear norm (MANN) to deal with The feature of were embedded into multi-core...
In this paper, we present a multimodal method based on densely connected convolution and bidirectional long-short-term-memory (BLSTM) for gesture recognition. The proposed learns spatial features of gestures through the convolutional network, then long-term temporal by BLSTM network. addition, fusion methods are evaluated our model, find that with different information can significantly improve recognition accuracy. This purely data driven approach achieves state-of-the-art accuracy ChaLearn...
The rapid development of artificial intelligence will profoundly change human social life. Based on this background, paper traces and studies the latest developments application results in field international intelligence, including achievements at technical level, characteristics market, trend competition pattern intelligence. For innovation planning AI, focus is outline AI + traditional industry, accelerating upgrading industrial building plateau, a safe convenient intelligent society,...
Cross project defect prediction is a valuable strategy for identifying software defects when historical data unavailable, yet selecting the right training predictive model challenging task. Additionally, dealing with numerous redundant features during significantly impacts accuracy. To tackle these challenges, this paper presents solution that combines candidate selection feature filtering. It Collaborative Filtering method Content Based to identify most appropriate project. For filtering,...
In this paper, we tackle the task of natural language video localization (NLVL): given an untrimmed and a description query, goal is to localize temporal segment within that best describes description. NLVL challenging at intersection understanding because may contain multiple segments interests describe complicated dependencies. Though existing approaches have achieved good performance, most them did not fully consider inherent differences between modalities. Here, propose Moment Relation...