A5005933485- ECG Monitoring and Analysis
- Gait Recognition and Analysis
- EEG and Brain-Computer Interfaces
- Advanced Computing and Algorithms
- Energy Load and Power Forecasting
- Interactive and Immersive Displays
- IoT and GPS-based Vehicle Safety Systems
- Video Surveillance and Tracking Methods
- Energy, Environment, Economic Growth
- Tactile and Sensory Interactions
- Stock Market Forecasting Methods
- Gastrointestinal Bleeding Diagnosis and Treatment
- Global Energy Security and Policy
- COVID-19 diagnosis using AI
- Human Pose and Action Recognition
- Economic and Technological Innovation
- Speech and Audio Processing
- Direction-of-Arrival Estimation Techniques
- Underwater Acoustics Research
- Complex Systems and Time Series Analysis
- Brain Tumor Detection and Classification
- Colorectal Cancer Screening and Detection
- Advanced Sensor and Energy Harvesting Materials
- Advanced Algorithms and Applications
- Forecasting Techniques and Applications
Tsinghua–Berkeley Shenzhen Institute
2023-2024
Tsinghua University
2023-2024
Chinese University of Hong Kong
2024
Queen's University
2023
National University of Defense Technology
2023
University Town of Shenzhen
2023
Qingdao Center of Resource Chemistry and New Materials
2020
Central University of Finance and Economics
2019
Modern Visual-Based Tactile Sensors (VBTSs) use cost-effective cameras to track elastomer deformation, but struggle with ambient light interference. Solutions typically involve using internal LEDs and blocking external light, thus adding complexity. Creating a VBTS resistant just camera an remains challenge. In this work, we introduce WStac, self-illuminating comprising mechanoluminescence (ML) whisker elastomer, camera, 3D printed parts. The ML inspired by the touch sensitivity of...
In response to the pressing need for robust disease diagnosis from gastrointestinal tract (GIT) endoscopic images, we proposed FLATer, a fast, lightweight, and highly accurate transformer-based model. FLATer consists of residual block, vision transformer module, spatial attention which concurrently focuses on local features global attention. It can leverage capabilities both convolutional neural networks (CNNs) transformers (ViT). We decomposed classification images into two subtasks: binary...
Arrhythmia automatic analysis techniques provide convenience for the prevention and diagnosis of cardiac disease. Aiming at classify abnormalities into 27 classes with either 12-lead, 6-lead, 4-lead, 3-lead or 2-lead multi-label ECG recordings, we develop a deep 1-Dimensional Convolutional Neural Network (1D CNN) residual block squeeze-and-excitation (SE) attention mechanism (namely 1D RANet). First, introduce SE CNN to extract features adaptively avoid vanishing gradient network...
This paper makes an event study on the Russo-Ukrainian War effect of stock return, including that S&P500, renewable energy industry, military and catering with a time window 29 days (39 for Teslas event). article will gain normal abnormal returns by using linear regression least square method. Then, hypothesis test be conducted Summation rate return (CAR). The result shown whether cumulative breaks confidence interval according to graph. Research results show even though war harms...
The traditional Direction of Arrival (DOA) estimation algorithms are based on model parameters, which depends the accuracy array model. When has errors, matching between and data will fail, affects performance to some extent. Therefore, this paper constructs nonlinear relationship received signal its spatial spectrum through neural network framework, uses data-driven deep learning realize DOA estimation. consists an autoencoder multiple parallel 1-D VGG networks achieve angle region....
Gait recognition aims to identify a person based on his unique walking pattern. Compared with silhouettes and skeletons, skinned multi-person linear (SMPL) models can simultaneously provide human pose shape information are robust viewpoint clothing variances. However, previous approaches have only considered SMPL parameters as whole yet explore their potential for gait thoroughly. To address this problem, we concentrate representations propose novel SMPL-based method named GaitSG...
Based on the knowledge of economics, this paper selects 22 macroeconomic indicators that best reflect overall economic situation United States. After differential, logarithmic and exponential preprocessing original data, paper, based power spectral analysis model, adaptively identifies periodicity selected indicators, visualize results. As a result, it screens out 11 with obvious periodicity. In process solving weighted distance principal component analysis, correlation test is first...