A5100437415- Advanced Graph Theory Research
- Graph Labeling and Dimension Problems
- graph theory and CDMA systems
- Face recognition and analysis
- Human Pose and Action Recognition
- Video Surveillance and Tracking Methods
- Face and Expression Recognition
- Computational Geometry and Mesh Generation
- Advanced Vision and Imaging
- Interconnection Networks and Systems
- Advanced Image and Video Retrieval Techniques
- Limits and Structures in Graph Theory
- Image Retrieval and Classification Techniques
- Advanced Neural Network Applications
- Remote-Sensing Image Classification
- Internet Traffic Analysis and Secure E-voting
- Advanced Software Engineering Methodologies
- Technology and Security Systems
- Robotics and Sensor-Based Localization
- Service-Oriented Architecture and Web Services
- Advanced Statistical Process Monitoring
- Target Tracking and Data Fusion in Sensor Networks
- Educational Reforms and Innovations
- Advanced Technologies in Various Fields
- Hand Gesture Recognition Systems
Tongji University
2024
Hebei University of Technology
2012-2023
Shanghai Key Laboratory of Trustworthy Computing
2023
East China Normal University
2023
Neijiang Normal University
2023
Northwestern Polytechnical University
2021-2022
Emilio Aguinaldo College
2022
NARI Group (China)
2019
South China University of Technology
2018
NYU Langone Health
2014
We present algorithms for recognizing human motion in monocular video sequences, based on discriminative conditional random field (CRF) and maximum entropy Markov models (MEMM). Existing approaches to this problem typically use generative (joint) structures like the hidden model (HMM). Therefore they have make simplifying, often unrealistic assumptions independence of observations given class labels cannot accommodate overlapping features or long term contextual dependencies observation...
We describe a mixture density propagation algorithm to estimate 3D human motion in monocular video sequences based on observations encoding the appearance of image silhouettes. Our approach is discriminative rather than generative, therefore it does not require probabilistic inversion predictive observation model. Instead, uses large capture data-base and computer graphics model order synthesize training pairs typical configurations together with their realistically rendered 2D These are...
Abstract Synthesis and re‐targeting of facial expressions is central to animation often involves significant manual work in order achieve realistic expressions, due the difficulty capturing high quality dynamic expression data. In this paper we address fundamental issues regarding use dense 3‐D data samples undergoing motions at video speeds, e.g. human expressions. utilize such for motion analysis re‐targeting, correspondences must be established between different frames same faces as well...
An important aspect to performing text categorization is selecting appropriate supervised classification and feature selection methods. A comprehensive benchmark needed inform best practices in this broad application field. Previous benchmarks have evaluated performance for a few methods limited ways optimize them. The present work updates prior by increasing the number of classifiers order magnitude, including adding recently developed, state‐of‐the‐art Specifically, study used 229 data...
This paper presents our joint research efforts on big data benchmarking with several industrial partners. Considering the complexity, diversity, workload churns, and rapid evolution of systems, we take an incremental approach in benchmarking. For first step, pay attention to search engines, which are most important domain Internet services terms number page views daily visitors. However, engine service providers treat data, applications, web access logs as business confidentiality, prevents...
Recently, rotation detection has gained much attention and shown its potential for accurate localization in remote sensing scenes. However, the objects images have a variety of directions, sizes, aspect ratios, which makes it difficult to locate classify objects. Therefore, object task is still facing great challenges field sensing. In this paper, we propose novel single-stage detector, includes feature alignment block (FAB), double regression branches (DRBs), circumcircle box (CRB). FAB...
Reliable 3D tracking is still a difficult task. Most parametrized deformable models rely on the accurate extraction of image features for updating their parameters, and are prone to failures when underlying feature distribution assumptions invalid. Active Shape Models (ASMs), other hand, based learning, thus require fewer reliable local than models, but fail easily they encounter situation which were not trained. In this paper, we develop an integrated framework that combines strengths both...
Process trace data (PTD) is an important type in semiconductor manufacturing and has a very large aggregate volume. While mining statistical analysis play key role the quality control of wafers, existence outliers adversely affects applications benefiting from PTD analysis. Due to complexities resultant outlier patterns, this paper proposes unified detection framework which takes advantages complexity reduction using entropy abrupt change cumulative sum (CUSUM) method. To meet practical...
The $book$ $embedding$ of a graph $G$ is to place the vertices on spine and draw edges pages so that in same page do not cross with each other. book embedding $matching$ if have maximum degree $1$. $thickness$ minimum number which graphs can be matching embedded. In this paper, we show thickness Cartesian product $K_p\Box C_q$of complete $K_p$ cycle $C_q$ equal $Δ(K_p\Box C_q)+1$.
The book embedding of a graph G is to arrange the set points on line (spine) and embed edges half-plane bounded by spine so that in same page do not intersect with each other. If maximum degree vertices 1, matching . thickness minimum number n can be embedded -page. In this paper, pseudo-Halin graphs determined.
Abstract The Proportional Hazards (PH) model is an important type of failure time regression which relates the occurrence probability critical failures to influential factors. However, little research work has been done on detecting changes in PH models fitted based different sets reliability data. This paper develops methods for change detection Cox models, also known as Semiparametric model, prediction and/or assessment time‐to‐failure data collected from subjects. effectiveness developed...
Microelectromechanical system (MEMS) accelerometers are small in size, low power consumption and easily integrated. They can be used intelligent hydraulic components to obtain the dynamic acceleration of monitor operating status system. In this study, based on noise characteristics MEMS accelerometer output signals, a wavelet modulus maxima denoising algorithm adaptive threshold estimation is proposed, which SureShrink choose right maxima. Then, signal-to-noise ratio mean-square-error as...
The matching book embedding of a graph G is to place its vertices on the spine, and arrange edges pages so that in same page do not intersect each other induced subgraphs are 1-regular. thickness minimum number required for any G, denoted by mbt(G). In this paper, generalized Petersen graphs determined.
The \emph{matching book embedding} of a graph $G$ is to arrange its vertices on the spine, and draw edges into pages so that every page do not intersect each other maximum degree one. thickness} minimum number in which can be matching embedded. In this paper, thickness Halin graphs determined.
In the domain of object detection, automatically selecting positive and negative samples methods have become a hot research topic in recent years. However, most them focus on improving sampling process but ignore relationship between size feature map, which shallow deep layers can capture small large objects well respectively. this paper, we propose multi-scale sample selection based statistical characteristics for detection. To improve robustness Intersection over Union (IoU) threshold,...
The Alon-Tarsi number of a graph $G$ is the smallest $k$ for which there an orientation $D$ with max outdegree $k-1$ such that Eulerian subgraphs even edges differs from odd edges. In this paper, we obtain Halin equals 4 when it wheel order and 3 otherwise.
In this paper, we give the classification of circulant graphs $C(\mathbb{Z}_{n},S)$ with $|S|=2$ and completely solve dispersability $C(\mathbb{Z}_{n},\{1, k\})$.
Highway alignment design is a nonlinear multi-objective optimization problem. Due to the uncertainty and complexity of constraints, it challenging develop suitable model for 3D road design. To comprehensively consider constraints realize intelligent design, this study puts forward novel highway optimize horizontal vertical in which length utility cost structure construction are projected into space. Finally, numerical solution carried out by MOPSO on MATLAB platform. The results show that...