A5102007341- Soft Robotics and Applications
- Advanced Neural Network Applications
- Robot Manipulation and Learning
- Hand Gesture Recognition Systems
- Advanced Vision and Imaging
- Safety Warnings and Signage
- Human Pose and Action Recognition
- Space Satellite Systems and Control
- Video Analysis and Summarization
- COVID-19 diagnosis using AI
- Robotic Path Planning Algorithms
- Visual Attention and Saliency Detection
- Vehicular Ad Hoc Networks (VANETs)
- Advanced Image and Video Retrieval Techniques
- Video Surveillance and Tracking Methods
- Hydrocarbon exploration and reservoir analysis
- Robotics and Sensor-Based Localization
- Web Data Mining and Analysis
- Multimodal Machine Learning Applications
- Computer Graphics and Visualization Techniques
- Robotic Locomotion and Control
- Autonomous Vehicle Technology and Safety
- Planetary Science and Exploration
- Mobile Ad Hoc Networks
- Domain Adaptation and Few-Shot Learning
Chinese University of Hong Kong, Shenzhen
2018-2025
Wannan Medical College
2024
Second Hospital of Anhui Medical University
2024
Anhui Medical University
2024
First Affiliated Hospital of Wannan Medical College
2024
Beijing Jiaotong University
2024
Research Institute of Petroleum Exploration and Development
2024
Kunming University of Science and Technology
2023
Sun Yat-sen University
2023
North China Electric Power University
2009-2023
Using the financial data of listed Chinese companies, we study impact COVID-19 on corporate performance. We show that has a negative firm The performance is more pronounced when firm's investment scale or sales revenue smaller. show, in an additional analysis, serious-impact areas and industries. These findings are among first empirical evidence association between pandemic
Dynamic hand gesture recognition is a challenging problem in the area of hand-based human–robot interaction (HRI), such as issues complex environment and dynamic perception. In context this problem, we learn from principle data-glove-based method propose based on 3D pose estimation. This uses estimation, data fusion deep neural network to improve accuracy gestures. First, 2D estimation OpenPose improved obtain fast method. Second, weighted sum utilized combine RGB, depth skeleton Finally,...
Abstract To deploy deep neural networks to edge devices with limited computation and storage costs, model compression is necessary for the application of learning. Pruning, as a traditional way compression, seeks reduce parameters weights. However, when network pruned, accuracy will significantly decrease. The decrease loss fine-tuning. When over many are pruned network’s capacity reduced heavily cannot recover high accuracy. In this paper, we apply knowledge distillation strategy abate...
Global temporal information and local semantic are essential cues for high-performance online object detection in videos. However, despite their promising accuracy most cases, state-of-the-art approaches have following two limitations: invalid background/scale suppression inadequate mining between frames. Many jobs currently focus on learning based a single frame. In this article, we propose an attentional global–local network; is one of the first attempts to fully use both types Attention...
New challenges such as automation, connection, electrification, and sharing (ACES) have brought disruptive changes to vehicles, transportation, mobility services, which urgently requires an ideal solution for sustainable transportation. This paper introduces the Internet a paradigm and, first time, proposes Transportation (TI), inspired by similarity between Referring construction ideas of Internet, this establishes framework TI, transportation router based on switching routing models,...
Aiming at the problems of accurate and fast hand gesture detection teleoperation mapping in hand-based visual dexterous robots, an efficient framework based on deep learning is proposed this article. It can achieve robots anchor-free network architecture by using RGB-D camera. First, early-fusion method HSV space proposed, effectively reducing background interference enhancing information. Second, a classification (HandClasNet) to realize localization detecting center corner points hands,...
How do humans recognize an action or interaction in the real world? Due to diversity of viewing perspectives, it is a challenge for identify regular activity when they observe from uncommon perspective. We argue that discriminative spatiotemporal information remains essential cue human recognition. Most existing skeleton-based methods learn optimal representation based on human-crafted criterion requires many labeled data and much effort. This article introduces adaptive neural networks...
Semantic Scene Completion (SSC) aims to reconstruct complete 3D scenes with precise voxel-wise semantics from the single-view incomplete input data, a crucial but highly challenging problem for scene understanding. Although SSC has seen significant progress due introduction of 2D semantic priors in recent years, occluded parts, especially rear-view scenes, are still poorly completed and segmented. To ameliorate this issue, we propose novel deep learning framework SSC, named Planar...
The COVID-19 pandemic continues to pose unprecedented threats and challenges global public health. Hospital Clinical Laboratory health institutions have been playing an important role in case detection, epidemic research decision-making, prevention control.To explore the current situation influencing factors of work stress medical workers hospital clinical laboratory fighting against COVID-19.A cluster random sampling method was used select seven hospitals from 14 tertiary Xiamen, selected...
In this paper, we propose a navigation algorithm oriented to multi-agent environment. This is expressed as hierarchical framework that contains Hidden Markov Model (HMM) and Deep Reinforcement Learning (DRL) structure. For simplification, term our method Hierarchical Navigation Network (HNRN). high-level architecture, train an HMM evaluate the agents perception obtain score. According score, adaptive control action will be chosen. While in low-level two sub-systems are introduced, one...
Visual hand-based robot teleoperation provides a powerful guarantee for robots to complete complex tasks. However, detection and distinction of dual hands on images are difficult because the small differences between left right hands. To solve this problem, parallel dual-hand method that combines features with relationship body pose is proposed achieve robust accurate detection. This includes hand module, estimation fusion module. In detector realizes fast by detecting center corner points...