A5037835022- Indoor and Outdoor Localization Technologies
- Speech and Audio Processing
- Underwater Vehicles and Communication Systems
- Energy Efficient Wireless Sensor Networks
- Millimeter-Wave Propagation and Modeling
- Wireless Networks and Protocols
- Hand Gesture Recognition Systems
- Advanced SAR Imaging Techniques
- Gait Recognition and Analysis
- Robotics and Sensor-Based Localization
- Radio Wave Propagation Studies
- Sparse and Compressive Sensing Techniques
- Non-Invasive Vital Sign Monitoring
- Target Tracking and Data Fusion in Sensor Networks
- Direction-of-Arrival Estimation Techniques
- Radiomics and Machine Learning in Medical Imaging
- Video Surveillance and Tracking Methods
- Inertial Sensor and Navigation
- GNSS positioning and interference
- AI in cancer detection
- Radar Systems and Signal Processing
- Advanced MIMO Systems Optimization
- Anomaly Detection Techniques and Applications
- Distributed Sensor Networks and Detection Algorithms
- Bluetooth and Wireless Communication Technologies
Chongqing University of Posts and Telecommunications
2016-2025
Shanghai Artificial Intelligence Laboratory
2025
Beijing Academy of Artificial Intelligence
2025
Chongqing Jiaotong University
2011-2024
China Mobile (China)
2022-2023
Rutgers, The State University of New Jersey
2021-2023
Princeton University
2023
Rutgers Sexual and Reproductive Health and Rights
2021
Jiangsu University of Science and Technology
2021
University of Tsukuba
2019-2020
Recently, hand gesture recognition systems have become increasingly interesting to researchers in the field of human-computer interfaces. Real-world for human dynamic is challenging as: 1) system must be robust various conditions; 2) there a rich diversity how people perform gestures, making difficult; and 3) detect recognize gestures continuously using unsegmented input streams order avoid noticeable lag between performing its classification. In this paper, address these challenges, we...
In recent years, non-contact radar detection technology has been able to achieve long-term and long-range for the breathing heartbeat signals. Compared with contact-based methods, it brings a more comfortable faster experience human body, gradually received attention in field of sensing. Therefore, this paper extends application millimeter-wave health care. The first transmits frequency-modulated continuous wave (FMCW) collects echo signals body. Then, phase information intermediate...
This paper uses an information-theoretic lens to view the error bound of wireless local area network (WLAN) localization, which is recognized as one superior candidate localization techniques in GPS-denied environment. Interestingly, we analogize process WLAN into information propagation a parallel Gaussian noisy channel, and then derive corresponding from channel capacity analogical system. Experimental results show that compared with widely-known Cramer-Rao Lower Bound (CRLB), proposed...
With the rapid increase of industrial systems, spectrum is stepping into era big data, and at same time resources are facing serious shortage. Cognitive system (CIS) based on cognitive radio can improve utilization by accessing idle licensed to primary user. However, CIS must find enough channels performing sensing. In this article, a reinforcement learning-based multislot double-threshold sensing with Bayesian fusion proposed sense which required faster while guaranteeing performance....
Abstract The demand for navigating pedestrian by using a hand-held mobile device increased remarkably over the past few years, especially in GPS-denied scenario. We propose new dead reckoning (PDR)-based navigation algorithm magnetic, angular rate, and gravity (MARG) sensors which are equipped existing commercial smartphone. Our proposed consists of step detection, stride length estimation, heading estimation. To eliminate gauge errors random bouncing motions, we designed reliable detection....
With the broad deployment of Wi-Fi networks, Received Signal Strength (RSS) based indoor localization has attained much interest both academia and industry. At present, most currently available techniques focus on increasing accuracy. However, few them take into account diversity signal distributions measurement error associated with RSS values owing to complicated environment, which consequently results in low robustness systems. Thus, motivation tackle this gripping problem, we design a...
The Internet of Things (IoT) has gradually changed the way people's lives due to its ability connecting everything together, and meanwhile accurate location sensing plays a crucial role in achieving this goal. Up now, as one most representative outdoor localization systems, global positioning system been widely used, but performance may be dramatically declined indoor environment serious multipath effect signal attenuation caused by complicated structure. At same time, fingerprint-based...
Channel state information (CSI) can provide phase and amplitude of multichannel subcarrier to better describe signal propagation characteristics. Therefore, CSI has become one the most commonly used features in indoor Wi-Fi localization. In addition, compared geometric localization method, fingerprint method advantages easy implementation high accuracy. However, as scale database increases, training cost processing complexity fingerprints will also greatly increase. Based on this, this...
The security threats caused by the popularity of Unmanned Aerial Vehicles (UAVs) have received much attention. In this paper, a UAV detection and identification system based on WiFi signal radio frequency (RF) fingerprint is proposed. firstly conducts after detected, fractal dimension (FD), axially integrated bispectra (AIB), square (SIB) are extracted as RF fingerprints due to their applicability reliability. Furthermore, we propose weighted AIB SIB identify UAVs. Since high dimensionality...
In recent years, a large number of researchers are endeavoring to develop wireless sensing and related applications as Wi-Fi devices become ubiquitous. As significant research branch, gesture recognition has one the hotspots. this paper, we propose WiCatch, novel device free system which utilizes channel state information recognize motion hands. First all, with aim catching weak signals reflected from hands, data fusion-based interference elimination algorithm is proposed diminish caused by...
The indoor localization systems based on wireless local area network received signal strength (RSS) have been widely applied due to the simplicity of system deployment as well easy implementation various mobile devices like smartphones. However, they are often suffered by major drawback extensive effort for location fingerprinting which is significantly labor-intensive and time-consuming. In response this compelling problem, we design an improved manifold alignment approach construct a...
Aiming at the problems of noise impact on parametric image hand gestures, difficulty gesture feature extraction, and inefficient utilization continuous time sequential information, we propose a inflated 3 dimensions (TS-I3D) convolutional neural network approach for recognition based frequency modulated wave (FMCW) radar sensor. Specifically, FMCW is used to acquire data, range speed in each frame signal are calculated by 2 fast Fourier transform. Then, range-Doppler map (RDM) generated...
In this article, a novel method for continuous hand gesture detection and recognition is proposed based on frequency modulated wave (FMCW) radar. Firstly, we adopt the 2-Dimensional Fast Fourier Transform (2D-FFT) to estimate range Doppler parameters of raw data, construct range-time map (RTM) Doppler-time (DTM). Meanwhile, apply Multiple Signal Classification (MUSIC) algorithm calculate angle angle-time (ATM). Secondly, segment gestures using decision threshold. Thirdly, central...
With the in-depth development of Internet Things (IoT) and constant discussion 6G visions, future indoor location-based services (LBSs) in 6G-enabled IoT is attracting people's attention. In fact, due to continuous updating network techniques, complexity environment number connected wireless access points (APs) will increase dramatically, which leads diversity received signal strength (RSS) on performance propagation distance estimation, resulting low positioning accuracy poor robustness....
Pedestrians represent agile and low-observable targets, especially under adverse weather conditions, whose trajectory tracking plays a crucial role in determining pedestrian behavior autonomous urban driving. Thus, this article presents millimeter-wave radar-based trajectory-tracking (MRPT) system that enables all-weather perception with high precision. More specifically, to improve the performance presence of strong background clutters, track-before-detection-based algorithm is proposed...
The cognitive industrial Internet of Things (CIIoT) can improve transmission performance by utilizing the spectrum licensed to a primary user (PU), providing that normal communication PU is not disturbed. However, traditional access schemes for CIIoT are difficult adapt various environments. In this article, <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$Q$</tex-math></inline-formula> -learning-based...
In recent years, unmanned aerial vehicles (UAVs) have received growing attention due to security threats issues. Though UAV detection and positioning systems are commonly used in various scenarios, most of the available still suffering from low accuracy susceptibility environment. Therefore, it is necessary design a highly accurate, versatile system. this paper, system based on multi-dimensional signal features proposed. The first step monitor communication channel state information (CSI)...
This paper presents a novel framework for recognition of facial action unit (AU) combinations by viewing the classification as sparse representation problem. Based on this framework, we represent image exhibiting combination AUs linear basis constituting an overcomplete dictionary. We build dictionary whose main elements are mean Gabor features AU under examination. The other randomly sampled from distribution (e.g., Gaussian distribution) that guarantees signal recovery. Afterwards, solving...
With the wide deployment of Wi-Fi networks, based indoor localization systems that are deployed without any special hardware have caught significant attention and become a currently practical technology. At same time, Magnetic, Angular Rate, Gravity (MARG) sensors installed in commercial mobile devices can achieve highly-accurate short time. Based on this, we design novel system by using built-in MARG module. The innovative contributions this paper include enhanced Pedestrian Dead Reckoning...
Wireless local area network (WLAN) fingerprinting has been extensively studied for indoor localization due to the formidable deployability of WLAN in environment, but one major bottleneck its practical implementation is extensive calibration effort required construct radio map through which time consuming and labor intensive. In response this compelling problem, we newly design graph-based semi-supervised manifold alignment approach relies on concept graph construction a cost-efficient with...
Recently, although radar sensors have been widely applied for hand gesture recognition (HGR) tasks, conventional radar-based HGR systems still two major challenges. First, these rely on supervised learning approaches to learn features, which normally require a large-scale labeled dataset address the overfitting problem. However, acquisition of such is time-consuming. Second, signature movement often influenced by micromotion caused other body parts, leads distorted motion resulting in poor...
An indoor target intrusion sensing technique has been used in many fields, such as smart home management, security monitoring, counter-terrorism, and disaster relief. At the same time, with wide deployment of wireless local area network (WLAN) general support IEEE 802.11 protocol by various mobile devices, can be realized based on existing WLAN infrastructure without requiring to carry any special device. However, approaches usually depend radio map construction huge labor time cost, which...
Digital whole-slide images are a unique way to assess the spatial context of cancer microenvironment. Exploring these characteristics will enable us better identify cross-level molecular markers that could deepen our understanding biology and related patient outcomes.We proposed graph neural network approach emphasises spatialisation tumour tiles towards comprehensive evaluation predicting profiles genetic mutations, copy number alterations, functional protein expressions from images. We...