A5032419801- Advanced Neural Network Applications
- Video Surveillance and Tracking Methods
- Advanced Image and Video Retrieval Techniques
- Geometric and Algebraic Topology
- Image Retrieval and Classification Techniques
- Advanced Image Processing Techniques
- Handwritten Text Recognition Techniques
- Text and Document Classification Technologies
- Metallurgical Processes and Thermodynamics
- Robotics and Sensor-Based Localization
- Mathematical Dynamics and Fractals
- Infrared Target Detection Methodologies
- Face and Expression Recognition
- Remote-Sensing Image Classification
- Image and Signal Denoising Methods
- Advanced Vision and Imaging
- Vehicle License Plate Recognition
- Gyrotron and Vacuum Electronics Research
- Medical Image Segmentation Techniques
- Advanced Topology and Set Theory
- Advanced Operator Algebra Research
- Infrastructure Maintenance and Monitoring
- Image Processing and 3D Reconstruction
- Fire Detection and Safety Systems
- Plasma Diagnostics and Applications
Zhangzhou Normal University
2015-2025
Inner Mongolia Agricultural University
2024
Peking University
2022
Auckland University of Technology
2017
Wuhan Institute of Technology
2015
Harbin University of Science and Technology
2011
Université de Lille
2010-2011
Hunan University
2010-2011
Deep multi-view clustering utilizes neural networks to extract the potential peculiarities of complementarity and consistency information among features. This can obtain a consistent representation that improves performance. Although multitude deep approaches have been proposed, most lack theoretic interpretability while maintaining advantages good In this paper, we propose an effective differentiable network with alternating iterative optimization for co-clustering termed bi-sparse (DBMC)...
Dimensionality reduction is commonly used to preprocess high-dimensional data, which an essential step in machine learning and data mining. An outstanding low-dimensional feature can improve the efficiency of subsequent tasks. However, existing methods dimensionality mostly involve datasets with sufficient labels fail achieve effective vectors for insufficient labels. In this paper, unsupervised multiple layered sparse autoencoder model studied. Its advantage that it reduces reconstruction...
For investigating the mechanism of high power microwave flashover and breakdown on dielectric surface with outgassing, firstly, theoretical modeling is put forward, including dynamic equations, particle-in-cell (PIC) method, secondary emission, Monte-Carlo collision (MCC) method outgassing model. Secondly, based modeling, 1D3V PIC-MCC code programmed by authors. By using this code, weak strong course under different gas moving velocities are studied numerically. The numerical results...
In this paper, we prove that separable subgroups have bounded packing in ambient groups. The notion was introduced by Hruska and Wise, and, particular, our result confirms a conjecture of theirs which states each subgroup virtually polycyclic group has the property.
In this paper, an electromagnetic-field FDTD method coupled with plasma fluid model is put forward to investigate the different physical phenomena of high power microwave (HPM) flashover and breakdown on inner outer surface output-window. Based above theoretical models, a one-dimensional (1D) electromagnetic field interaction code programmed by authors. By using code, HPM output-window are simulated. The numerical results could be concluded as follows. For surface, output pulse shortened...
In this paper, we establish that, for statistically convex-cocompact actions, contracting elements are exponentially generic in counting measure. Among others, the following exponential genericity results obtained as corollaries set of hyperbolic relatively groups, rank-1 CAT(0) and pseudo-Anosov mapping class groups. Regarding a proper action, non-contracting is proven to be growth-negligible. particular, sufficiently large subgroup, provided that subgroup has purely growth. By Roblin's...