A5047373940- Indoor and Outdoor Localization Technologies
- Speech and Audio Processing
- Wireless Networks and Protocols
- Advanced Adaptive Filtering Techniques
- Context-Aware Activity Recognition Systems
- Underwater Vehicles and Communication Systems
- Wireless Power Transfer Systems
- Non-Invasive Vital Sign Monitoring
- Energy Harvesting in Wireless Networks
- Advanced MIMO Systems Optimization
- Gait Recognition and Analysis
- Mathematical Analysis and Transform Methods
- Energy Efficient Wireless Sensor Networks
University of Science and Technology Beijing
2024-2025
Guangdong University of Technology
2021
Human activity recognition (HAR) is a key technology in the field of human–computer interaction. Unlike systems using sensors or special devices, WiFi channel state information (CSI)-based HAR are noncontact and low cost, but they limited by high computational complexity poor cross-domain generalization performance. In order to address above problems, reconstructed CSI tensor deep learning based lightweight system (Wisor-DL) proposed, which firstly reconstructs signals with sparse signal...
Gesture recognition is an essential part in the field of human–computer interaction (HCI) and Internet Things system. Compared with existing technologies based on wearable sensors dedicated devices, approaches using WiFi channel state information (CSI) signals are more desirable for passive fine-grained gesture recognition. However, CSI-based systems usually suffer from high model complexity low accuracy caused by environmental dynamics. To address these issues, we propose a robust system...
Recent studies have shown that WiFi channel state information (CSI) based approaches for human activity recognition (HAR) is successful. However, the performance of these often deteriorate when deployed to a new industrial environment. To solve this problem without retraining, we present novel CSI tensor cross-domain HAR approach (TensFi). Specifically, activity-related first separated from original through an ensemble empirical mode decomposition (EEMD) algorithm. Then, sparse signal...
With the widespread use of WiFi devices and availability channel state information (CSI), CSI-based device-free localization (DFL) has attracted lots attention. Fingerprint-based methods are primary solutions for DFL, but they faced with fingerprint similarity problem due to complex environment low bandwidth commercial WiFi. Meanwhile, fingerprints may change unpredictably multipath signal propagation in time-varying environments. To tackle these problems, this paper proposes an adaptive...