A5100441306- Advanced Graph Neural Networks
- Hand Gesture Recognition Systems
- Complex Network Analysis Techniques
- Human Pose and Action Recognition
- Text and Document Classification Technologies
- Spectroscopy and Chemometric Analyses
- Hearing Impairment and Communication
- Tactile and Sensory Interactions
- Diffusion and Search Dynamics
- Point processes and geometric inequalities
- Gait Recognition and Analysis
- Human Mobility and Location-Based Analysis
- Advanced Computational Techniques and Applications
- Blind Source Separation Techniques
- Image and Signal Denoising Methods
- Cryptography and Data Security
- Adversarial Robustness in Machine Learning
- Power Systems and Technologies
- Interactive and Immersive Displays
- Privacy-Preserving Technologies in Data
Shanghai Jiao Tong University
2024
PLA Information Engineering University
2019-2020
System Equipment (China)
2019
Massachusetts Institute of Technology
2012-2014
Tianjin University of Technology and Education
2014
Analyzing the rich information behind heterogeneous networks through network representation learning methods is signifcant for many application tasks such as link prediction, node classifcation and similarity research. As evolve over times, interactions among nodes in make exhibit dynamic characteristics. However, almost all existing focus on static which ignore In this paper, we propose a novel approach DHNE to learn representations of networks. The key idea our construct comprehensive...
Our real-time continuous gesture recognition system addresses problems that have previously been neglected: handling both gestures are characterized by distinct paths and hand poses; determining how when the should respond to gestures. probabilistic framework based on hidden Markov models (HMMs) unifies of two forms Using information from states in HMM, we can identify different phases: pre-stroke, nucleus post-stroke phases. This allows appropriately require a discrete response those...
We developed a gesture salience based hand tracking method, and spotting recognition method on concatenated hidden Markov models. A 3-fold cross validation using the ChAirGest development data set with 10 users gives an F1 score of 0.907 accurate temporal segmentation rate (ATSR) 0.923. The average final is 0.9116. Compared joint position from Kinect SDK, our 3.7% absolute increase in score.
This article proposes a novel scheme, SupRTE, to suppress backdoor injection in federated learning via robust trust evaluation, which effectively prevents malicious updates from infiltrating the model aggregation process. The evaluation process SupRTE consists of two components: 1) behavior representation extractor, creating individual profiles for each client through multidimensional information; 2) scorer, measuring discrepancies between and benign clients as scores by utilizing grading...
I propose a systematic hierarchical approach to continuous gesture analysis using unifying framework based on abstract hidden Markov models (AHMMs). With this framework, will develop gesture-based interactive interface that allows users do both manipulative and communicative gestures without artificial restrictions, hence enabling natural interaction.
Mining the rich structure and semantic information hidden in heterogeneous networks is one of important tasks network representation learning. At present, there are relatively few studies on for which contains different types nodes link relationships. Most science based homogeneous where objects same entity type links between also type. In this paper, we propose a new learning method network. The core tow parts: First, according to meta-paths, get weights attention mechanism. And then,...
This article analyzes the characteristics of complete sets electrical appliances enterprise production and current situation demand information application, points out that platform should be based on product data management, including CAX/CAPP/PDM/ERP integrated application platform. In this architecture, integration management is responsible for design process, generate BOM process BOM, form basis ERP data, order original input, logistics cash flow. Using corresponding interface design,...
Network representation learning, which aims to learn the low-dimensional representations of vertices, has attracted considerable research efforts recently. Most existing network learning methods mainly focus on static networks, extract and condense information without temporal information. However, in real world, networks keep evolving evolution process contains a lot important But it is difficult quantify node structure dynamic networks. To address this problem, we propose method, call as...