A5064335604- Face and Expression Recognition
- Image Retrieval and Classification Techniques
- Biometric Identification and Security
- Reinforcement Learning in Robotics
- Advanced Text Analysis Techniques
- Machine Learning and ELM
- Face recognition and analysis
- Topic Modeling
- Generative Adversarial Networks and Image Synthesis
- Adversarial Robustness in Machine Learning
- Advanced Image and Video Retrieval Techniques
- Bayesian Modeling and Causal Inference
- Neural Networks and Applications
- Advanced Image Processing Techniques
- Advanced Image Fusion Techniques
- Natural Language Processing Techniques
- Sentiment Analysis and Opinion Mining
- Bayesian Methods and Mixture Models
- Image and Signal Denoising Methods
- Multimodal Machine Learning Applications
- Text and Document Classification Technologies
- Remote-Sensing Image Classification
- Advanced Neural Network Applications
- Gaussian Processes and Bayesian Inference
- Explainable Artificial Intelligence (XAI)
Tianjin University of Science and Technology
2014-2025
Beijing University of Technology
2023
Fuzhou University
2023
First Affiliated Hospital of University of South China
2021
University of South China
2021
Xidian University
2021
Fudan University
2021
University of Science and Technology Beijing
2012
Tianjin University
2008-2009
Finger vein recognition is a biometric technology using finger veins to authenticate person, and due its high degree of uniqueness, liveness, safety, it widely used.The traditional Symmetric Local Graph Structure (SLGS) method only considers the relationship between image pixels as dominating set, uses relevant theories tap features.In order better extract features, taking into account location information direction image, this paper presents novel feature extraction method,...
Multimodal biometrics combine a variety of biological features to have significant impact on identification performance, which is newly developed trend in technology. This study proposes novel multimodal recognition model based the stacked extreme learning machines (ELMs) and canonical correlation analysis (CCA) methods. The model, has symmetric structure, found high potential for biometrics. works as follows. First, it learns hidden-layer representation images using layer by layer. Second,...
Local Graph Structure (LGS) and its variation Symmetric (SLGS) have been proven to be effective for image recognition. However, they shortcomings without considering the contribution of difference between target pixel surrounding pixels, pixels feature value pixel. To overcome traditional methods, this paper proposes a Difference (DSLGS) algorithm finger vein The DSLGS operator considers different pixel, making extracted more stable. experiment results show that proposed has better...
Abstract In this article, the containment control problem is studied for hybrid multi‐agent systems (MASs), which comprised of continuous‐time and discrete‐time dynamic agents. For first‐order MASs, two effective distributed protocols are designed when followers have dynamics, one protocol dynamics. Meanwhile, second‐order we also propose three kinds to solve control. By utilizing stability theory system transformation method, some criteria derived solving MASs. Simulation examples provided...
Wireless sensor networks (WSNs) usually consist of a large number power-constrained sensors deployed to collect information in specific sensed field. How gather data an energy-efficient manner is always very important issue WSNs. Recently, cluster-based aggregation widely used reduce energy consumption and prolong network lifetime. However, most existing schemes proposed fail consider the factor quality service (QoS), especially source-to-sink delay data-loss probability. To address this...
The two-stage linear discrimination analysis (TSLDA) is a feature extraction technique to solve the small size sample problem in field of image recognition.The TSLDA has retained all subspace information between-class scatter and within-class scatter.However, four subspaces may not be entirely beneficial for classification, regularization procedure eliminating singular metrics higher time complexity.In order address these drawbacks, this paper proposes an improved discriminant (Improved...
Text matching is a critical task in natural language processing to measure semantic similarity between two texts. A significant portion of online texts are labeled with variety coarse topic responses. These supervised indicators can provide prior structured and explicable semantics for textual modeling. However, most existing state-of-the-art neural network methods cannot benefit from such complementary signals. Therefore, we propose novel Topic Supervision BERT-based model (TSB) text...
Rotational symmetry is important for many applications in computer graphics, vision, and image processing. However, it remains difficult to design an effective algorithm automatic recognition. In this paper, we present a rotational detection algorithm, which easy use can determine both the center radius of supporting region without human interaction. Our derived from frieze-expansions approach improved through radius-based expansion idea. Multiresolution pyramid used accelerate process. We...