Abstract:In recent years, due to advances in informatization, the national economy has become more dependent on networks. As a result, the network and information security has become a prominent problem for the national security and social stability. Driven by economic interests and the game between countries reflected in the growing cyberspace confrontations, network attacks with high-tech, high concealment and long-term sustainability become one of the major challenges in the network security. In this paper, the certain kind of covert attack was referred as the evasive network attack (ENA). Firstly, the main characteristics of ENA and the challenges it brings in current security systems were analyzed, based on which the latest developments at home and abroad for ENA detection and other related studies were reviewed then. Finally, key technical issues and future research directions in this field were summarized.

Abstract:Interactions of ET-1 (Endothelin-1) with its receptors ETA (Endothelin-A) and ETB (Endothelin-B) regulate the vascular tone, maintain the blood pressure, and are closely related to cardiovascular diseases. Binding of ET-1 with ETB in the vascular endothelium induces vasodilation, while binding with ETA and ETB in vascular smooth muscle results in vasoconstriction. Because ET-1 only induces vasodilation in vivo when the blood flow is present, we speculate that the shear stress of the blood flow may control the conformation of ET-1 through its structural flexibility, thus regulate its binding with ETB. By flow molecular dynamics simulation, the conformational changes of ET-1 in uniform flow were studied with its center of mass constrained. It is found that the C-terminal of ET-1 gets closer to the N-terminal in the simulation, resulting in a compact structure. This finding may provide guidance for the study on the interaction between ET-1 and ETB and the design of ET-1-based cardiovascular drugs.

Abstract:In this paper, the convergence time required to achieve consensus of dynamic systems was studied under the uniform averaging model. In each time step, a node’s value was updated to some weighted average of its neighbors’ and its old values. The case was studied when the underlying network was dynamic. Our analysis results show that dynamic networks exhibit fast convergence behavior as long as the nodes’ degrees change gradually, even under very mild connectivity assumptions.

Abstract:Imbalanced data exist widely in the real world and their classification is a hot topic in the field of machine learning. A clustering-based enhanced AdaBoost algorithm was proposed to improve the poor classification performance produced by the traditional algorithm in classifying the minority class of imbalanced datasets. The algorithm firstly constructs balanced training sets by the clustering-based undersampling, using K-means clustering to cluster the majority class and extract cluster centroids and then merge with all minority class instances to generate a new balanced training set. To avoid the declining of the classification accuracy caused by the shortage of training sets owing to too few minority class samples, SMOTE (Synthetic Minority Oversampling Technique) combining the clustering-based undersampling was used. Next, the misclassification loss function in the basic classifier of the AdaBoost algorithm was modified based on the costsensitive learning theory to assign asymmetric misclassification losses to samples of different classes. The experimental results show that, the proposed algorithm makes the model training samples more representative and greatly increases the classification accuracy of the minority class, keeping the overall classification performance.

Abstract:In order to solve the huge data problem for high resolution or ultra-high resolution video, the International Organization for Standardization in video coding has been developing the most advanced video compression standard – High Efficiency Video Coding (HEVC), namely H.265. As an important part of this standard, the complexity and performance of intra-coding will seriously affect the compression complexity and performance of the standard. To reduce the complexity of HEVC intra-coding, an algorithm of re-determining the candidate mode list was presented is this paper. In the proposed algorithm, according to the different modes of the first index in candidate mode list, an adaptive method was utilized to determine which mode will be kept in candidate mode list, then the number of modes which are needed to perform rate distortion optimization was reduced. Experimental results show that, compared with original HM8.0, the proposed algorithm saves 24.50% on average in time for intra-coding with the same compression performance. The proposed algorithm can be combined with other level of fast coding optimization to further reduce the complexity.

Abstract:The weather has a profound influence on human’s daily life and the weather forecasting has always been a topic of great concern. With the economic development and social progress, people’s requirements for daily weather forecasting has become higher and higher. Information provided by the general circulation models (GCMs) can describe well some of the weather parameters at a large scale, but GCMs fail to provide detailed weather information at a regional or local scale for impact assessment studies. Outputs from GCMs are usually of low spatial resolutions. A common approach to bridge the scale mismatch is downscaling. In the present study, two methods, i.e., the statistical multiple linear regression and the BP neural network, were proposed to downscale large scale reanalysis data to daily temperature extremums at a local point, Shenzhen national meteorological station. The data used in this study are NCEP/NCAR (National Centers for Environmental Prediction/National Centre for Atmospheric Research) reanalysis dataset for the 2000~2012 period and daily observations of maximum temperature and minimum temperature at Shenzhen station for the same period. The two methods were compared in this study. Results show that both methods can simulate well the daily temperature extremums at Shenzhen station, but the performance of the statistical downscaling method is more stable than the BP neural network.

Abstract:The route planning engine has already become an important part for an online map system. The route planning algorithm is the key for the engine. The existing improvements for A* algorithm are mainly on the preprocessing part in which the roadmap data were layered statically. In this paper, an adaptive hierarchical method was proposed with an improved heuristic function which has goal-direction process. It greatly improves the efficiency and usability of A* algorithm in the engineering road planning system. The experiment result shows that the algorithm takes up only 42% of the search space and 13% of the search time when compared with the general A* algorithm.

Abstract:Traditional GVP (geometry-preserving visual phrases) image retrieval algorithm is not suitable for handling the large-scale image retrieval because of its high time complexity. In this paper, FSF-GVP (frequency statistics featuregeometry- preserving visual phrases) algorithm, which combined word frequency statistic characteristics and GVP algorithm, was proposed. FSF-GVP algorithm counts visual word frequency characteristics of an image to be searched and image database to get similar result set and dissimilar result set. Then FSF-GVP algorithm uses the GVP algorithm to sort the similar result set, which improves the retrieval efficiency. The experiment results on Oxford 5K dataset show that FSFGVP is suitable for the large-scale real-time image retrieval on the premise of ensuring the accuracy of retrieving result and improving the retrieval efficiency.

Abstract:The image registration is a process of establishing spatial correspondences between two images. It is widely used in the computer vision, the remote sensing data analysis and the image processing. Especially in the image-guided radiation therapy, the image registration plays an important role. Recently, the scale-invariant feature transform (SIFT) has been used in the medical image registration, and obtained promising results. However, SIFT is apt to detect blob features which cannot reflect properly motions of lungs. In this paper, a hybrid feature detection method, which can detect lung tissue features effectively based on Harris and SIFT algorithms, was proposed. In addition, a novel method which can remove mismatched landmarks was also proposed. A series of thoracic CT images were tested by using the proposed algorithm. The quantitative and qualitative evaluations show that our method is much better than the conventional SIFT method especially in the case of large deformations of lungs during the respiration.

Abstract:The freehand 3D ultrasound is a new imaging tool based on the ultrasonic imaging mechanism and spatial positioning apparatus. It can be used to image all kinds of samples and in-vivo human subjects freely by integrating the spatial positioning senor on the conventional 2D probe of the 2D ultrasound system. In this paper, a freehand 3D ultrasound with the application of minimally invasive surgery was developed. The popular distance-weighted interpolation method was well studied and applied in our imaging system. Experimental results on both phantom and chicken kidney demonstrate that the developed freehand 3D ultrasound system can produce satisfied images to prove the feasibility of the system.