检索结果-南通市图书馆

A Novel Simple Visual Tracking Algorithm Based on Hashing and Deep Learning

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2017年第5期26卷 1073-1078页

作者： ZHU Suguo du junping REN Nan Beijing Key Laboratory of Intelligent Telecommunication Software and Multimedia School of Computer ScienceBeijing University of Posts and Telecommunications

Deep network has been proven efficient and robust to capture object features in some *** still remains in the stage of classifying or detecting *** the field of visual tracking,deep network has not been applied *** of the reasons is that its time consuming made the strong method could not meet the speed need of visual tracking.A novel simple tracker is proposed to complete tracking task.A simple six-layer feed-forward backpropagation neural network is applied to capture object ***,this representation is not robust enough when illumination changes or drastic scale changes in dynamic *** improve the performance and not to increase much time spent,image perceptual hashing method is employed,which extracts low frequency information of object as its fingerprint to recognize the object from its structure.64-bit characters are calculated by it,and they are utilized to be the bias terms of the neutral *** leads more significant improvement for performance of extracting sufficient object *** we take particle filter to complete the tracking process with the proposed *** experimental results demonstrate that the proposed algorithm is efficient and robust compared with the state-of-the-art tracking methods.

关键词： Visual tracking Image perceptual hashing Deep learning Neural network

Cross-media Hot Topic Auto-tracking Model Based on Semantics and Temporal Context

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2015年第3期24卷 529-534页

作者： LIANG Meiyu du junping ZHOU Yipeng Beijing Key Lab of Intelligent Telecommunication Software and Multimedia School of Computer ScienceBeijing University of Posts and Telecommunications School of Computer and Information Engineering Beijing Technology and Business University

As the Internet multimedia information grows explosively, seeking an automatic technology to realize the effective organization and management of crossmedia emergency information is significantly necessary. A novel cross-media hot topic auto-tracking model based on semantics and temporal context is proposed in this paper. According to the semantic correlations of cross-media information, we learn the image visual semantics by the text semantics based on the Latent Dirichlet Allocation probability model, and establish the unified cross-media information description on the same semantic level. Also a semantics-based two-step feature dimension-reduction scheme is proposed to establish the efficient semantic feature space. The self-adaptive learning of topic model is realized to track the dynamic changes in the topic. Experimental results demonstrate that the proposed method outperforms the existing methods, which further improves the effect of hot topic auto-tracking.

关键词： Cross-media data mining Topic autotracking Semantic model Latent dirichlet allocation(LDA) Temporal context

Crowd Density Field Estimation Based on Crowd Dynamics Theory and Social Force Model

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2019年第3期28卷 521-528页

作者： WEI Xinlei du junping LIANG Meiyu XUE Zhe Beijing Key Lab of Intelligent Telecommunication Software and Multimedia School of Computer ScienceBeijing University of Posts and Telecommunications

The local crowd density and the crowd distribution estimation tasks are useful but *** are two main problems which limit the performance of existing algorithms. The first one is that there are not enough labeled training samples to build the highperformance estimation model. Another one is that existing methods lack the supports of physical theories of the crowd. To remedy them, a novel crowd density field model is proposed, which is deduced by jointing crowd dynamics theory and social force model. A crowd counting method based on the proposed crowd density field model is introduced to measure the proposed crowd density field model. Extensive experiments confirm the effectiveness of the proposed plan.

关键词： Social force model Crowd dynamics Crowd density field Computer vision

Self-adaptive spatial image denoising model based on scale correlation and SURE-LET in the nonsubsampled contourlet transform domain

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2014年第9期57卷 72-86页

作者： LIANG MeiYu du junping LIU HongGang Beijing Key Laboratory of Intelligent Telecommunication Software and Multimedia School of Computer ScienceBeijing University of Posts and Telecommunications

A novel self-adaptive image denoising model based on scale correlation and Stein’s unbiased risk estimate-linear expansion of thresholds(SURE-LET)in the nonsubsampled contourlet transform domain is proposed in this *** we implement the multidimensional and translation invariant decomposition for spatial images by the nonsubsampled contourlet transform,and establish the image cross-scale description *** combining the scale correlation,we make improvements for the existing SURE-LET denoising idea and establish the self-adaptive denoising *** scale correlation calculation is needed for the coefficients at different scales and sub-bands to determine whether the coefficients are retained or processed with the adaptive SURE-LET threshold *** meanwhile a new local context self-adaptive threshold strategy is proposed in the process of scale correlation *** results both on spatial images and standard images demonstrate that the proposed algorithm performs significantly better in terms of both the visual subjective evaluation and the quantitative objective *** method can achieve better noise suppression,and effectively retain image edge details.

关键词： spatial images image denoising nonsubsampled contourlet transform SURE-LET scale correlation

The Browsing Pattern and Review Model of Online Consumers Based on Large Data Analysis

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2015年第1期24卷 58-64页

作者： NING Lianju WANG Haoyu FENG Xin du junping School of Economics and Management Beijing University of Posts and Telecommunications Beijing Key Lab of Intelligent Telecommunication Software and Multimedia School of Computer ScienceBeijing University of Posts and Telecommunications

In the context of online shopping, commodity information and consumer reviews are main factors that will affect purchasing behavior. Started from the preference of commodity information browsing and the inherent property of online reviews, this paper focuses on the browsing data for statistical analysis and the interval distribution of consumer reviews based on the real data of360 buy which is the domestic large-scale B2 C commerce website in China. Researches find that commodity information browsing time distribution on Internet is fragment and can be depicted by the fat tail effect. It also demonstrated that user’s browsing patterns are related to the type of information and the displaying of the information,which means that the pieces of the picture and the length of the titles affecting the clicks rate. Reviews on the interval distribution can be depicted by the power-law function and there is a monotonically increasing relationship between power-exponent and the customers’ concerns with the corresponding commodity, the higher the exponent,the higher the degree of consumer attention. The finding obtained some basic rules of the browsing mode and review model, which is of important significance for future research.

关键词： Interval distribution Online consumer Browsing pattern Reviews model Trustworthiness.

Hierarchical-Based Object Detection with Improved Locality Sparse Coding

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2016年第2期25卷 290-295页

作者： ZHU Suguo du junping REN Nan LIANG Meiyu Beijing Key Laboratory of Intelligent Telecommunication software and Multimedia School of Computer ScienceBeijing University of Posts and Telecommunications

This paper proposes to extend the hierarchical method to be adapted to sequential frames, aiming at detecting the moving object in dynamic scenes. A novel two-layer model is proposed, in which dictionaries are learned through three different stages and the locality constrained sparse representation is improved. This leads more significant improvement for performance of both static image classification and moving object detection. The experimental results demonstrate that the proposed algorithm is efficient and robust compared with the state-of-the-art classification methods, and also able to detect moving object in the sequential frames accurately.

关键词： Hierarchical detection Sparse coding Classification Object detection

Adaptive multiple video sensors fusion based on decentralized Kalman filter and sensor confidence

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2017年第6期60卷 133-144页

作者： Qingping LI junping du Suguo ZHU Liang XU Beijing Key Laboratory of Intelligent Telecommunication Software and Multimedia School of Computer ScienceBeijing University of Posts and Telecommunications

The fusion of multiple video sensors provides an effective way to improve the robustness and accuracy of video surveillance systems. In this paper, an adaptive fusion method based on a decentralized Kalman filter(DKF) and sensor confidence is presented for the fusion of multiple video sensors. The adaptive scheme is one of the approaches used for preventing the divergence problem of the filter when statistical values of the measurement noises of the system models are not available. By introducing the sensor confidence, we can adaptively adjust the measurement noise covariance matrix of the local DKFs and thus, determine the weight of each sensor more correctly in the fusion procedure. Also, the DKF applied here can make full use of redundant tracking data from multiple video sensors and give more accurate fusion results in an efficient manner. Finally,the fusion result with improved accuracy is obtained. Experimental results show that the proposed adaptive decentralized Kalman filter fusion(ADKFF) method works well in the case of real-world video sequences and exhibits more promising performance than single sensors and comparative fusion methods.

关键词： video sensors fusion decentralized Kalman filter target tracking sensor confidence video surveillance

FDBST: Fast Discovery of Bursty Spatial-Temporal Topic

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2020年第1期29卷 168-176页

作者： ZHU Chuangying du junping ZHANG Qiang SHI Lei Lee JangMyung Beijing Key Laboratory of Intelligent Telecommunication Software and Multimedia School of Computer ScienceBeijing University of Posts and Telecommunications Research Institute of Computer Information and Communication Pusan National University

Discovering the hot topic among trends is an essential way to manage public opinion. But the dynamic property makes it a tough task as the existing methods of detection process are time-consuming. With the new Fast discovery of burst spatial-temporal topic(FDBST) method, the uptime of topic discovery is kept within one second and there is no sacrificing of topic quality, the noisy topics are all kept off, the time-varying problem and the sparsity problem are easily tackled by using spatial-temporal characteristics. How the FDBST works? It is triggered by a burst term based on social data trends, while the data is generated within a small time interval in a region while the irrelevant data is excluded during this process. By fusing the regional topics, the potential burst topic is *** experiments show the preferable effects of the FDBST and it is an outperforms state-of-the-art approaches in terms of effectiveness.

关键词： Social network Real-time Spatialtemporal Burst topic Data stream

Construction of Ni–Al–Ru EAM potential and application in misfit dislocation system

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Progress in Natural Science:Materials International 2020年第4期30卷 539-544页

作者： Enlai Yue Tao Yu Chongyu Wang junping du Central Iron and Steel Research Institute Department of Physics Tsinghua University Department of Mechanical Science and Bioengineering Osaka University

In Ni-based single crystal superalloys, ruthenium is sometimes introduced as one of the creep resistances through retarding the thermally-activated deformation processes, such as dislocation glide and climb. In the present study, an embedded-atom-method potential of Ni–Al–Ru system was constructed. Using the present potential,the effect of Ru on the lattice misfit between γ(Ni) and γ′(Ni3Al) phases was investigated. The results show that Ru doping decreases the lattice misfit, which is consistent with the calculations from first-principles. The interaction between Ru and misfit dislocation was also studied. The results of interaction energy between Ru and misfit dislocation, and the simulations of misfit dislocation motion in the Ru-doping system by molecular dynamics method both show that Ru possesses the pinning effect on the misfit dislocation. These results can provide useful help for further insights of the role of Ru in the alloys.

关键词： Embedded-atom-method potential Ruthenium Ni–Al–Ru Misfit dislocation