A5015980929- Human Pose and Action Recognition
- Gait Recognition and Analysis
- Caching and Content Delivery
- Anomaly Detection Techniques and Applications
- Peer-to-Peer Network Technologies
- Advanced Neural Network Applications
- Image Enhancement Techniques
- Advanced Image Fusion Techniques
- Electrolyte and hormonal disorders
- Parathyroid Disorders and Treatments
- AI in cancer detection
- Image Processing and 3D Reconstruction
- Energy Harvesting in Wireless Networks
- Advanced Image Processing Techniques
- Opportunistic and Delay-Tolerant Networks
- Dialysis and Renal Disease Management
- Internet Traffic Analysis and Secure E-voting
- Advanced MIMO Systems Optimization
- Video Surveillance and Tracking Methods
- Context-Aware Activity Recognition Systems
- Hand Gesture Recognition Systems
- Millimeter-Wave Propagation and Modeling
Northwestern Polytechnical University
2024
Xi'an Polytechnic University
2023-2024
ShanghaiTech University
2018
Guilin University of Electronic Technology
2012
To fully leverage the labeled action data, one-shot learning solutions are designed for skeleton-based human recognition recently. These tend to employ deep metric methods enlarge inter-class distances while suppressing intra-class variations in embedding space. However, a notable issue is that actions even from same class may exhibit different characteristics since individual differences among performers. Consequently, hinder model's ability generalize unseen classes. alleviate this...
Abstract Skeleton‐based human action recognition is gaining significant attention and finding widespread application in various fields, such as virtual reality human‐computer interaction systems. Recent studies have highlighted the effectiveness of graph convolutional network (GCN) based methods this task, leading to a remarkable improvement prediction accuracy. However, most GCN‐based overlook varying contributions self, centripetal centrifugal subsets. Besides, only single‐scale temporal...
To tackle the uplink pilot contamination problem in massive multiple-input multiple-output (MIMO) systems, current researches only relied on angle of arrival at base station. However, this information is insufficient when users share same scattering environment. In paper, we propose a novel strategy by exploiting user mobility. Due to limited scatterers around users, first investigate channel sparsity and derive corresponding angle-Doppler frequency domain power spectrum (AD-CPS). We then...