A5101587813- Evacuation and Crowd Dynamics
- Digital Transformation in Industry
- COVID-19 epidemiological studies
- Music and Audio Processing
- Infrastructure Resilience and Vulnerability Analysis
- Underwater Vehicles and Communication Systems
- Underwater Acoustics Research
- Speech and Audio Processing
- Engineering Education and Technology
- Advanced Multi-Objective Optimization Algorithms
- Maritime Navigation and Safety
- Scheduling and Timetabling Solutions
- Metaheuristic Optimization Algorithms Research
- Economic and Technological Systems Analysis
- Music Technology and Sound Studies
Guilin University of Electronic Technology
2022-2023
Zhejiang University of Technology
2021-2022
Ministry of Education of the People's Republic of China
2022
Polyphonic sound event detection (SED) is the task of detecting time stamps and class that occurred during a recording. Real life events overlap in recordings, their durations vary dramatically, making them even harder to recognize. In this paper, we propose Convolutional Recurrent Neural Networks (CRNNs) extract hidden state feature representations; then, self-attention mechanism using symmetric score function introduced memorize long-range dependencies features CRNNs extract. Furthermore,...
The recognition of underwater acoustic targets plays a crucial role in marine vessel monitoring. However, traditional target models suffer from limitations, including low accuracy and slow prediction speed. To address these challenges, this article introduces novel approach called the Multi-Gradient Flow Global Feature Enhancement Network (MGFGNet) for automatic targets. Firstly, new spectrogram feature fusion scheme is presented, effectively capturing both physical brain-inspired features...
As the complexity of real-world engineering problems increases, numerous efficient constraint-handling methods and optimization algorithms have emerged recently. However, majority research consider method algorithm independently. In this paper, we propose a constraint-based mutation operator, in which constraint violation objective function are considered simultaneously. We define pbest individuals as best top 5% violators if all infeasible. way, could guide population move towards feasible...
The traditional agent-based model requires high computing power of the central processing unit. Thus, an improved combined with discrete event simulation method is proposed. result equation-based Susceptible-Exposed-Infective-Asymptomatic-Recovered (SEIAR) same parameter combination, which has been demonstrated to be effective, used verify validity this model. Additionally, analysis based on results Contact Tracing Measure (CTM), Location-Based Checking-Testing (LCTM), Lockdown (LM), Mobile...