Chinese University of HongKong

HongKong, China

Chinese University of HongKong

HongKong, China

Time filter

Source Type

Chen L.,Tsinghua University | Chen L.,Chinese University of Hongkong | Chen X.,Beijing University of Posts and Telecommunications | Zhang C.,Southern Utah University
Nonlinear Analysis: Real World Applications | Year: 2011

The bipolar EulerPoisson system in physics consists of the conservation laws for the electron and ion densities and their current densities, coupled with the Poisson equation for the electrostatic potential. The limit of vanishing ratio of the electron mass to the ion mass in the n-dimensional flat torus is proved in the case of well prepared initial data. The limiting system is composed of two separated equations, where the equation for electron is the incompressible Euler equation with damping, which means physically that the evolution for electrons and ions can be treated as separated motions in the small ratio case. © 2010 Elsevier Ltd. All rights reserved.


PubMed | Chinese University of Hongkong and Sun Yat Sen University
Type: Journal Article | Journal: Metabolic brain disease | Year: 2015

Hypertension is considered one of the most important controllable risk factors for white matter lesion (WML). Our previous work found that stroke-prone renovascular hypertensive rats (RHRSP) displayed a high rate of WML. This study aimed to investigate the WML in RHRSP from MRI, pathology and behavior. RHRSP model was established by two-kidney, two-clipmethod and kept for 20 weeks. WML was decteted by magnetic resonance imaging (MRI) and loyez staining. Cognition was tested by morris water maze (MWM). Vascular changes were observed by HE staining on brain and carotid sections. Ultrastucture of blood brain barrier (BBB) were observed by transmission electron microscope. Immunofluorescence was used to detect albumin leakage and cell proliferation. T(2)-weighted MRI scans of RHRSP displayed diffuse, confluent white-matter hyperintensities. Pathological examination of the same rat showed marked vacuoles, disappearence of myelin and nerve fibers in white matter, supporting the neuroimaging findings. Spatial learning and memory impairment were observed in RHRSP. The small arteries in brain exhibited fibrinoid necrosis, hyalinosis and vascular remodeling. BBB disruption and plasma albumin leakage into vascular wall was observed in RHRSP. Increased cell proliferation in subventricular zone was seen in RHRSP. RHRSP demonstrated spontaneous WML and cognitive impairment. Hypertensive small vessel lesions and BBB disruption might paly causative factors for the onset and development of WML. The characteristic features of WML in RHRSP suggested it a valid animal model for WML.


PubMed | Chinese University of Hongkong and Sun Yat Sen University
Type: Journal Article | Journal: Parasitology research | Year: 2016

Sj16 is a Schistosoma japonicum-derived protein (16kDa in molecular weight) that has been identified as an immune modulation molecule, but the mechanisms of modulation of immune responses are not known. In this report, we aimed to investigate the host immune regulation mechanism by recombinant Sj16 (rSj16) and thus illuminate the molecular mechanism of immune evasion by S. japonicum. The effect of rSj16 and rSj16 mutants on the biology of dendritic cells (DCs) was assessed by examining DC maturation, cytokine production, and expression of surface markers by flow cytometry and enzyme-linked immunosorbent assay. We found that rSj16 significantly stimulated interleukin (IL)-10 production and inhibited LPS-induced bone marrow-derived dendrite cell (BMDC) maturation in a dose-dependent manner. By using antibody neutralization experiments and IL-10-deficient (knockout) mice, we confirmed that the inhibitory effect of rSj16 on LPS-induced BMDCs is due to its induction of IL-10 production. To understand how rSj16 induces the production of IL-10, we analyzed the protein sequence and revealed two potential nuclear localization signals (NLS) in Sj16. The N-terminal NLS (NLS1) is both necessary and sufficient for translocation of rSj16 to the nucleus of BMDCs and is important for subsequent induction of IL-10 production and the inhibition of BMDC maturation by rSj16. The results of our study concluded that the ability of rSj16 to inhibit DC functions is IL-10 dependent which is operated by IL-10R signal pathway. This study also confirmed that NLS is an important domain associated with increased production of IL-10. Our findings will extend the current understanding on host-schistosome relationship and provide insight about bottleneck of parasitic control.


Liang Y.,Macau University of Science and Technology | Chai H.,Macau University of Science and Technology | Liu X.-Y.,Macau University of Science and Technology | Xu Z.-B.,Xi'an Jiaotong University | And 2 more authors.
BMC Medical Genomics | Year: 2016

Background: One of the most important objectives of the clinical cancer research is to diagnose cancer more accurately based on the patients' gene expression profiles. Both Cox proportional hazards model (Cox) and accelerated failure time model (AFT) have been widely adopted to the high risk and low risk classification or survival time prediction for the patients' clinical treatment. Nevertheless, two main dilemmas limit the accuracy of these prediction methods. One is that the small sample size and censored data remain a bottleneck for training robust and accurate Cox classification model. In addition to that, similar phenotype tumours and prognoses are actually completely different diseases at the genotype and molecular level. Thus, the utility of the AFT model for the survival time prediction is limited when such biological differences of the diseases have not been previously identified. Methods: To try to overcome these two main dilemmas, we proposed a novel semi-supervised learning method based on the Cox and AFT models to accurately predict the treatment risk and the survival time of the patients. Moreover, we adopted the efficient L1/2 regularization approach in the semi-supervised learning method to select the relevant genes, which are significantly associated with the disease. Results: The results of the simulation experiments show that the semi-supervised learning model can significant improve the predictive performance of Cox and AFT models in survival analysis. The proposed procedures have been successfully applied to four real microarray gene expression and artificial evaluation datasets. Conclusions: The advantages of our proposed semi-supervised learning method include: 1) significantly increase the available training samples from censored data; 2) high capability for identifying the survival risk classes of patient in Cox model; 3) high predictive accuracy for patients' survival time in AFT model; 4) strong capability of the relevant biomarker selection. Consequently, our proposed semi-supervised learning model is one more appropriate tool for survival analysis in clinical cancer research. © 2016 Liang et al.


Liu L.,National University of Defense Technology | Fieguth P.,University of Waterloo | Wang X.,Chinese University of HongKong | Pietikainen M.,University of Oulu | Hu D.,National University of Defense Technology
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2016

In recent years, a wide variety of different texture descriptors has been proposed, including many LBP variants. New types of descriptors based on multistage convolutional networks and deep learning have also emerged. In different papers the performance comparison of the proposed methods to earlier approaches is mainly done with some well-known texture datasets, with differing classifiers and testing protocols, and often not using the best sets of parameter values and multiple scales for the comparative methods. Very important aspects such as computational complexity and effects of poor image quality are often neglected. In this paper, we propose a new extensive benchmark (RoTeB) for measuring the robustness of texture operators against different classification challenges, including changes in rotation, scale, illumination, viewpoint, number of classes, different types of image degradation, and computational complexity. Fourteen datasets from the eight most commonly used texture sources are used in the benchmark. An extensive evaluation of the recent most promising LBP variants and some non-LBP descriptors based on deep convolutional networks is carried out. The best overall performance is obtained for the Median Robust Extended Local Binary Pattern (MRELBP) feature. For textures with very large appearance variations, Fisher vector pooling of deep Convolutional Neural Networks is clearly the best, but at the cost of very high computational complexity. The sensitivity to image degradations and computational complexity are among the key problems for most of the methods considered. © Springer International Publishing AG 2016.


Wang L.,CAS Shenzhen Institutes of Advanced Technology | Chen H.,CAS Institute of Software | Li S.,CAS Shenzhen Institutes of Advanced Technology | Meng H.M.,CAS Shenzhen Institutes of Advanced Technology | Meng H.M.,Chinese University of HongKong
Speech Communication | Year: 2012

Speech visualization is extended to use animated talking heads for computer assisted pronunciation training. In this paper, we design a data-driven 3D talking head system for articulatory animations with synthesized articulator dynamics at the phoneme level. A database of AG500 EMA-recordings of three-dimensional articulatory movements is proposed to explore the distinctions of producing the sounds. Visual synthesis methods are then investigated, including a phoneme-based articulatory model with a modified blending method. A commonly used HMM-based synthesis is also performed with a Maximum Likelihood Parameter Generation algorithm for smoothing. The 3D articulators are then controlled by synthesized articulatory movements, to illustrate both internal and external motions. Experimental results have shown the performances of visual synthesis methods by root mean square errors. A perception test is then presented to evaluate the 3D animations, where a word identification accuracy is 91.6% among 286 tests, and an average realism score is 3.5 (1 = bad to 5 = excellent). © 2012 Elsevier B.V. All rights reserved.


Yang Z.,University of Tokyo | Yu J.X.,Chinese University of Hongkong | Liu Z.,Chinese University of Hongkong | Kitsuregawa M.,University of Tokyo
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2010

Discovery of evolving regions in large graphs is an important issue because it is the basis of many applications such as spam websites detection in the Web, community lifecycle exploration in social networks, and so forth. In this paper, we aim to study a new problem, which explores the evolution process between two historic snapshots of an evolving graph. A formal definition of this problem is presented. The evolution process is simulated as a fire propagation scenario based on the Forest Fire Model (FFM) [17]. We propose two efficient solutions to tackle the issue which are grounded on the probabilistic guarantee. The experimental results show that our solutions are efficient with regard to the performance and effective on the well fitness of the major characteristics of evolving graphs. © Springer-Verlag Berlin Heidelberg 2010.


Zhang C.,Nanjing University of Aeronautics and Astronautics | Guo P.,Chinese University of HongKong | Ehmann K.F.,Northwestern University | Li Y.,Nanjing University of Aeronautics and Astronautics
Materials and Manufacturing Processes | Year: 2015

Microgroove, as a form of surface texturing, has a wide array of industrial applications. However, the use of conventional methods to machine microgrooves leads to a number of problems including large burrs, high cutting forces, and poor machining quality. In this paper, ultrasonic elliptical vibration cutting is used to assist microgrooves turning on cylindrical workpiece surfaces. The elliptical locus in the cutting process is generated by a newly designed 2D resonant ultrasonic vibrator. A series of microgrooves cutting experiments without and with the ultrasonic elliptical vibration-assistance is performed to verify the effects of the ultrasonic elliptical vibrations as compared to the ordinary cutting method. The generated cutting forces, burr suppression action, and microgroove surface quality are compared for the two classes of processes. Comparison results show the effectiveness of elliptical vibration-assisted microgroove cutting in reducing cutting forces and improving microgrooves machining quality for difficult-to-cut materials. The results also show that ultrasonic elliptical vibration-assisted cutting improves the microgroove turning process with respect to cutting forces, microgroove surface roughness, and burr formation for difficult-to-cut materials. © Taylor & Francis Group,LsLC.


Liu W.,CAS Shenzhen Institutes of Advanced Technology | Hu C.,CAS Shenzhen Institutes of Advanced Technology | He Q.,CAS Shenzhen Institutes of Advanced Technology | Meng M.Q.-H.,Chinese University of HongKong
2010 IEEE International Conference on Information and Automation, ICIA 2010 | Year: 2010

In order to achieve the goal of high accuracy and low cost in a visual localization system, we present a novel localization method based on four inexpensive video cameras. The method mainly consists of two parts: The "16-points interpolation algorithm" is proposed to enhance the accuracy of 2D coordinates of the detected target on the image plane. Another important aspect is that the Perpendicular Foot Method (PFM) is used to calculate the 3D coordinates of the target. Simulation and real experimental results show that the stability of image coordinates (x, y) is significantly improved by the "16-points interpolation algorithm", and the PFM algorithm is better than the traditional LSM (Least Square Method) algorithm. The localization accuracy of this system can reach 5mm, when the target is moved along with the y axes direction which is perpendicular to the optical axes of CCD cameras. ©2010 IEEE.


Wei M.,CAS Shenzhen Institutes of Advanced Technology | Wei M.,Chinese University of HongKong | Li Y.,Nanjing Normal University | Wu J.,CAS Shenzhen Institutes of Advanced Technology | And 2 more authors.
Proceedings - 2011 International Conference on Cyberworlds, Cyberworlds 2011 | Year: 2011

We present a rapid and effective point simplification algorithm for surface reconstruction which can represent different levels-of-detail. The core of this algorithm is to generate an approximately minimal set of adaptive balls covering the whole surface by defining and minimizing local quadric error functions. First, the feature points are extracted by simple thresholding curvatures, Second, for the non-feature points, they are covered by distinct balls. The size of each ball varies and reflects how curved the local surface is. Once the size of radius is fixed, the points in each ball will be substituted by an optimized point. Thus, the simplified surface consists of extracted feature points and optimized points. we can employ this algorithm to produce coarse-to-fine models by controlling a general error level, and name it as ESimp for short. Worthy of note, the error level of each ball may be adaptively adjusted according to the local curvature and density of the center of this ball which can avoid holes generation. Finally, the simplified points are triangulated by Cocone algorithm. This algorithm has been applied to a set of large scanned models. Experimental results demonstrate that it can generate high-quality surface approximation with feature preservation. © 2011 IEEE.

Loading Chinese University of HongKong collaborators
Loading Chinese University of HongKong collaborators