A5070678631- Sparse and Compressive Sensing Techniques
- Tensor decomposition and applications
- Image and Signal Denoising Methods
- Chinese history and philosophy
- Face and Expression Recognition
- Advanced Image Processing Techniques
- Advanced Neuroimaging Techniques and Applications
- Remote-Sensing Image Classification
- Advanced Image Fusion Techniques
- Traffic Prediction and Management Techniques
- Advanced Clustering Algorithms Research
- Blind Source Separation Techniques
- Data Management and Algorithms
- Big Data and Business Intelligence
- Video Surveillance and Tracking Methods
- Text and Document Classification Technologies
- Advanced SAR Imaging Techniques
- Aesthetic Perception and Analysis
- Cognitive Science and Education Research
- Statistical Methods and Inference
- Soil, Finite Element Methods
- Digital Media Forensic Detection
- Advanced Computing and Algorithms
- Neuroinflammation and Neurodegeneration Mechanisms
- Neural Networks and Applications
Southwest Jiaotong University
2023-2025
PowerChina (China)
2023-2024
Macau University of Science and Technology
2021-2023
China Power Engineering Consulting Group (China)
2023
Sun Yat-sen University
2022
First Affiliated Hospital of Zhengzhou University
2022
University of South China
2021-2022
University of Electronic Science and Technology of China
2018-2020
Guizhou University of Finance and Economics
2009-2012
Guizhou Electric Power Design and Research Institute
2010
Completing missing entries in multidimensional visual data is a typical ill-posed problem that requires appropriate exploitation of prior information the underlying data. Commonly used priors can be roughly categorized into three classes: global tensor low-rankness, local properties, and nonlocal self-similarity (NSS); most existing works utilize one or two them to implement completion. Naturally, there arises an interesting question: concurrently make use multiple unified way, such they...
Federated Learning (FL) provides a novel paradigm for privacy-preserving machine learning, enabling multiple clients to collaborate on model training without sharing private data. To handle multi-source heterogeneous data, Vertical (VFL) has been extensively investigated. However, in the context of VFL, label information tends be kept one authoritative client and is very limited. This poses two challenges VFL scenario. On hand, small number labels cannot guarantee train well with informative...
Most multi-view clustering methods based on shallow models are limited in sound nonlinear information perception capability, or fail to effectively exploit complementary hidden different views. To tackle these issues, we propose a novel Subspace-Contrastive Multi-View Clustering (SCMC) approach. Specifically, SCMC utilizes set of view-specific auto-encoders map the original data into compact features capturing its structures. Considering large semantic gap from modalities, project multiple...
Tensor analysis has received widespread attention in high-dimensional data learning. Unfortunately, the tensor are often accompanied by arbitrary signal corruptions, including missing entries and sparse noise. How to recover characteristics of corrupted make it compatible with downstream clustering task remains a challenging problem. In this article, we study generalized transformed low-rank representation (TTLRR) model for simultaneously recovering data. The core idea is find latent...
Abstract Face recognition (FR) systems based on convolutional neural networks have shown excellent performance in human face inference. However, some malicious users may exploit such powerful to identify others' images disclosed by victims' social network accounts, consequently obtaining private information. To address this emerging issue, synthesizing protection with visual and protective effects is essential. existing methods encounter three critical problems: poor effect, limited...
Deterministic wind power prediction can be used for long time-scale optimization of dispatching systems, but the probability and fluctuation range results cannot calculated. A Bayesian LSTM neural network (BNN-LSTM) is constructed based on networks by placing a priori distributions top layer weight parameters. First, temporal convolutional (TCNN) to process historical time-series data prediction, which extract correlation features learn trend changes data. Then, mutual information entropy...
Abstract The entorhinal cortex is of great importance in cognition and memory, its dysfunction causes a variety neurological diseases, particularly Alzheimer's disease (AD). Yet so far, research on still limited. Here, we provided the first single‐nucleus transcriptomic map primate aging. Our result revealed that synapse signaling, neurogenesis, cellular homeostasis, inflammation‐related genes pathways changed cell‐type‐specific manner with age. Moreover, among 7 identified cell types,...
With the arrival of era big data, data can no longer be expressed simply in two-dimensional form, and demand for value industry is getting higher higher, which promotes emergence many industries derived from data. Artificial intelligence one products this new era. Through continuous efforts scientists, artificial has developed to an unprecedented situation, but Turing Award winner Judy Bohr believes that current only weak intelligence, order achieve universal intelligent machines must learn...
Multi-view clustering (MVC) aims at exploiting the consistent features within different views to divide samples into clusters. Existing subspace-based MVC algorithms usually assume linear subspace structures and two-stage similarity matrix construction strategies, thereby posing challenges in imprecise low-dimensional representation inadequacy of exploring consistency. This paper presents a novel hierarchical for method via integration intra-sample, intra-view, inter-view learning models. In...