A5100323037- Robotics and Sensor-Based Localization
- Autonomous Vehicle Technology and Safety
- Advanced Memory and Neural Computing
- Advanced Neural Network Applications
- Video Surveillance and Tracking Methods
- Robotic Path Planning Algorithms
- Advanced Vision and Imaging
- Anomaly Detection Techniques and Applications
- Advanced Image and Video Retrieval Techniques
- CCD and CMOS Imaging Sensors
- Advanced Computational Techniques and Applications
- Domain Adaptation and Few-Shot Learning
- Robot Manipulation and Learning
- Advanced Algorithms and Applications
- Reinforcement Learning in Robotics
- Power Systems and Technologies
- 3D Shape Modeling and Analysis
- Smart Grid and Power Systems
- Industrial Technology and Control Systems
- Advanced Sensor and Control Systems
- Human Pose and Action Recognition
- Remote Sensing and LiDAR Applications
- 3D Surveying and Cultural Heritage
- Traffic control and management
- Neural dynamics and brain function
Tongji University
2018-2025
Shanghai Chengtou (China)
2019-2025
PLA Information Engineering University
2023-2024
Technical University of Munich
2014-2024
Zhejiang Institute of Mechanical and Electrical Engineering
2015-2024
Hangzhou City University
2024
Anhui University of Technology
2013-2024
Energy Research Institute
2018-2024
Jilin Agricultural University
2019-2024
Hebei University of Technology
2023-2024
Blockchain is the core technology used to create cryptocurrencies, like bitcoin. As part of fourth industrial revolution since invention steam engine, electricity, and information technology, blockchain has been applied in many areas such as finance, judiciary, commerce. The current paper focused on its potential educational applications explored how can be solve some education problems. This article first introduced features advantages following by exploring for education. Some innovative...
A combination of resistive switching and magnetic modulation gives rise to the integration room temperature ferromagnetism (spin) electrical properties (charge) into a simple Pt/Co:ZnO/Pt structure due formation oxygen vacancy-based conductive filaments. This is promising for broadening applications random access memories encode quaternary information. Detailed facts importance specialist readers are published as "Supporting Information". Such documents peer-reviewed, but not copy-edited or...
As a bio-inspired and emerging sensor, an event-based neuromorphic vision sensor has different working principle compared to the standard frame-based cameras, which leads promising properties of low energy consumption, latency, high dynamic range (HDR), temporal resolution. It poses paradigm shift sense perceive environment by capturing local pixel-level light intensity changes producing asynchronous event streams. Advanced technologies for visual sensing system autonomous vehicles from...
Although great progress has been made in generic object detection by advanced deep learning techniques, detecting small objects from images is still a difficult and challenging problem the field of computer vision due to limited size, less appearance, geometry cues, lack large-scale datasets targets. Improving performance wider significance many real-world applications, such as self-driving cars, unmanned aerial vehicles, robotics. In this article, first-ever survey recent studies...
To enhance the active safety performance for automated electric vehicles (AEVs) at driving limits, collaborative control of four-wheel steering (4WS) and direct yaw-moment (DYC) is adopted. deal with external disturbance modeling error, tube-based model predictive (MPC) applied to algorithm design, which takes improvement handling stability path-tracking into considerations. Taking constraints account, including vector constraints, lateral rollover prevention error integrated controller...
Follow-up study of coronavirus disease 2019 (COVID-19) survivors has rarely been reported. We aimed to investigate longitudinal changes in the characteristics COVID-19 after discharge.A total 594 discharged from Tongji Hospital Wuhan February 10 April 30, 2020 were included and followed up until May 17, 2021. Laboratory radiological findings, pulmonary function tests, electrocardiogram, symptoms signs analyzed.257 (51.2%) patients had at least one symptom 3 months post-discharge, which...
Robot learning through kinesthetic teaching is a promising way of cloning human behaviors, but it has its limits in the performance complex tasks with small amounts data, due to compounding errors. In order improve robustness and adaptability imitation learning, hierarchical strategy proposed: low-level comprises only behavioral supervised high-level constitutes policy improvement. First, Gaussian mixture model (GMM)-based dynamical system formulated encode motion from demonstration. We then...
Deep neural networks (DNNs) often perform poorly in the presence of domain shift and category shift. How to upcycle DNNs adapt them target task remains an important open problem. Unsupervised Domain Adaptation (UDA), especially recently proposed Source-free (SFDA), has become a promising technology address this issue. Nevertheless, existing SFDA methods require that source share same label space, consequently being only applicable vanilla closed-set setting. In paper, we take one step...
Learning-based methods have demonstrated clear advantages in controlling robot tasks, such as the information fusion abilities, strong robustness, and high accuracy. Meanwhile, on-board systems of robots limited computation energy resources, which are contradictory with state-of-the-art learning approaches. They either too lightweight to solve complex problems or heavyweight be used for mobile applications. On other hand, training spiking neural networks (SNNs) biological plausibility has...
Pedestrian detection has attracted enormous research attention in the field of Intelligent Transportation System (ITS) due to that pedestrians are most vulnerable traffic participants. So far, almost all pedestrian solutions based on conventional frame-based camera. However, they cannot perform very well scenarios with bad light condition and high-speed motion. In this work, a Dynamic Active Pixel Sensor (DAVIS), whose two channels concurrently output gray-scale frames asynchronous...
Drowsiness driving is a principal factor of many fatal traffic accidents. This paper presents the first event-based drowsiness detection (EDDD) system by using recently developed neuromorphic vision sensor. Compared with traditional frame-based cameras, sensors, such as Dynamic Vision Sensors (DVS), have high dynamic range and do not acquire full images at fixed frame rate but rather independent pixels that output intensity changes (called events) asynchronously time they occur. Since events...
DATA REPORT article Front. Neurorobot., 18 June 2019 Volume 13 - | https://doi.org/10.3389/fnbot.2019.00038
Predicting the future trajectories of surrounding agents has become an crucial problem to be solved for safety autonomous vehicles. Recent studies based on Long Short Term Memory (LSTM) networks have shown powerful abilities model social interactions. However, many these approaches focus spatial interactions neighborhood but ignore temporal that accompany In this paper, we propose a Hierarchical Spatio-Temporal Attention architecture (HSTA), which activates utilization with different...
Deep learning methods have achieved excellent results in the field of grasp detection. However, deep learning-based models for general object detection lack proper balance accuracy and inference speed, resulting poor performance real-time tasks. This work proposes an efficient network with n-channel images as inputs robotic grasp. The proposed is a lightweight generative structure one stage. Specifically, Gaussian kernel-based representation introduced to encode training samples, embodying...
Through vehicle-to-vehicle (V2V) communication, autonomizing a vehicle platoon can significantly reduce the distance between vehicles, thereby reducing air resistance and improving road traffic efficiency. The gradual maturation of control technology is enabling platoons to achieve basic driving functions, permitting large-scale scheduling planning, which essential for industrialized applications generates significant economic benefits. Scheduling planning are required in many aspects...
Most finger vein authentication systems suffer from the problem of small sample size. However, data augmentation can alleviate this to a certain extent but did not fundamentally solve category diversity. So researchers resort pre-training or multi-source joint training methods, these methods will lead user privacy leakage. In view above issues, paper proposes federated learning-based framework (FedFV) size and diversity while protecting privacy. Through under FedFV, each client share...