A5100613448- Advanced Neural Network Applications
- Advanced Vision and Imaging
- Human Pose and Action Recognition
- Per- and polyfluoroalkyl substances research
- Retinal Imaging and Analysis
- Optical measurement and interference techniques
- Natural Language Processing Techniques
- Topic Modeling
- Biodiesel Production and Applications
- Water Quality Monitoring and Analysis
- Anomaly Detection Techniques and Applications
- Neural Networks and Applications
- Medical Image Segmentation Techniques
- Toxic Organic Pollutants Impact
- Image and Object Detection Techniques
- Image Processing Techniques and Applications
- Nonlinear Dynamics and Pattern Formation
- Industrial Vision Systems and Defect Detection
- Domain Adaptation and Few-Shot Learning
- Image Enhancement Techniques
- Image and Video Stabilization
- Neural Networks Stability and Synchronization
- Teleoperation and Haptic Systems
- Video Surveillance and Tracking Methods
- Mathematical and Theoretical Epidemiology and Ecology Models
Craft Group (China)
2022-2024
Qinghai University
2023
Beijing University of Technology
2021
Ping An (China)
2021
State Key Joint Laboratory of Environment Simulation and Pollution Control
2021
Tsinghua University
2021
Beijing Jiaotong University
2006-2020
Northeast Forestry University
2018
Chongqing University
2016
Lanzhou University
2005-2013
Nowadays robot is supposed to demonstrate human-like perception, reasoning and behavior pattern in social or service application. However, most of the existing motion planning methods are incompatible with above requirement. A potential reason that navigation algorithms usually intend treat people as another kind obstacle, hardly take principle awareness into consideration. In this paper, we attempt model proxemics group blend it scenario perception robot. For purpose, a clustering method...
The detection of moving object is important in many tasks, such as video surveillance and tracking. Although there are some methods for the detection, it still a challenging area. In this paper, new method which combines Kirsch operator with Optical Flow (KOF) proposed. On one hand, used to compute contour objects video. other adopted establish motion vector field sequence. Then Otsu implemented after order distinguish background clearly. Finally information fuses label sequences. experiment...
Xuan Zhou, Xiao Zhang, Chenyang Tao, Junya Chen, Bing Xu, Wei Wang, Jing Xiao. Proceedings of the 2021 Conference North American Chapter Association for Computational Linguistics: Human Language Technologies. 2021.
A novel mechanism reduction method that integrates path flux analysis with multi generations and sensitivity (MPFASA) is proposed to reduce the complex detailed chemistry generate skeletal mechanisms. At first, (MPFA) used efficiently mechanisms; then (SA) applied further eliminate redundant species their related reactions on basis of obtained by MPFA method. Detailed mechanisms methane n-heptane are reduced MPFASA method, which validated in context autoignition perfectly stirred reactor for...
Various surface defects in automated fiber placement (AFP) processes affect the forming quality of components. In addition, defect detection usually requires manual observation with naked eye, which leads to low production efficiency. Therefore, automatic solutions for recognition have high economic potential. this paper, we propose a multi-scale AFP algorithm, named spatial pyramid feature fusion YOLOv5 channel attention (SPFFY-CA). The (SPFFY) adopts dilated convolutions (SPDCs) fuse maps...
This work investigated the roles of spontaneous synchronized rhythmic oscillations in information processing a globally coupled neural network. We found that encoding input aperiodic signals, there exists range optimal synchronous states network, where transmission rate and coding efficiency network are maximized. Our results indicate it is not weaker or stronger but an appropriate state may be more functional significance sensory encoding.
This paper presents a hybrid method using machine learning approach for named entity recognition (NER). A system built based on this is able to achieve reasonable performance with minimal training data and gazetteers. The differs from previous learning-based systems in that it uses maximum entropy model (MEM) hidden Markov (HMM) successively. We report the of our proposed NER British National Corpus (BNC). In process, we first use MEM identify entities corpus by imposing some temporary...
According to the characteristics of retinal fundus images, one novel method registration based on phase-correlation was proposed in this paper. This takes advantage optic disc match two images coarsely. Meanwhile edge detection applied extract abundant veins and arteries order raise precision registration. also overcomes gray-sensitive deficiency method. And experiment results show that not only enhances but reduces time complexity
Nowadays, the existing EOD robots usually are operated manually; it is very hard to operate. According theory of stereo vision, a new explosive ordnance disposal (EOD) robot with binocular vision system developed. The captures real-time images, makes camera calibrations Zhang's method, preprocesses matches feature point, then gets coordinate point suspicious objects, programming Matlab7.0 and pattern matching library EasyMatch machine software eVision, displays images on console, finally,...
A control system design of a new explosive ordnance disposal (EOD) robot based on the binocular stereo vision location is presented in this paper. The principle and motion are described. Real-time xPC target through Matlab/RTW method run successfully test prototype furnished with PC/104, satisfactory result obtained.
Image registration is an important task of image processing. In parallax mosaic, the geometric a global minimized residual in translation model. order to realize quick and robust image, coarse-to-fine strategy using coarse images weighted with detail proposed. Taking advantage multi-resolution characteristic Wavelet, original pyramid decomposed by Wavelet transformation. The result upper layer can be initial value next registration. better this way applied vertical bottom, correlative curve...
Abstract Aircraft coating process has been an important part in manufacturing of modern aviation products. For defect detection, the manual observation with naked eyes is usually utilized, which leads to low production efficiency. In this paper, authors propose improved YOLOv5‐OBB channel‐spatial attention block (CSAB), feature pyramid non‐local module (FPNM) and structured sparsity slimming criterion (SSSC). The CSAB can pay more effective channel information features from dimension target...
Fast and accurate evaluation of fish freshness has a scientific practical value for preservation, deposit further processing fish. Bio-impedance method, which was widely used to diagnose monitor diseases body in medical area, employed research on fast freshwater A bio-impedance measurement system constructed based LabVIEW with four electrodes sweep frequency method. The characteristics crucian investigated during its storage. result indicated that the impedance phase changed loss fish, when...
Epilepsy is a group of disorders characterized by recurrent paroxysmal electrical discharges the cerebral cortex that result in intermitted disturbances brain function. The damage induced seizure severe, so it significant to predict preictal state epileptic seizure. aim this work compare and estimate different nonlinear analysis methods predicting seizure, including approximate entropy, Lempel-Ziv complexity, spectral entropy C <sub xmlns:mml="http://www.w3.org/1998/Math/MathML"...