Modelling salient visual dynamics in videos

Automatic video annotation is a critical step for content-based video retrieval and browsing. Detecting the focus of interest in video frames automatically can benefit the tedious manual labeling process. However, producing an appropriate extent of visually salient regions in video sequences is a challenging task. Therefore, in this work, we propose a novel approach for modeling dynamic visual attention based on spatiotemporal analysis. Our model first detects salient points in three-dimensional video volumes, and then uses the points as seeds to search the extent of salient regions in a novel motion attention map. To determine the extent of attended regions, we use the maximum entropy in the spatial domain to analyze the dynamics derived by spatiotemporal analysis. Our experiment results show that the proposed dynamic visual attention model achieves high precision value of 70% and reveals its robustness in successive video volumes.     

交通流异常数据检测研究及实证

针对道路交通系统实时交通流数据普遍存在的异常现象,提出一种基于最小二乘支持向量机的交通异常数据检测方法。运用相空间重构技术,将时间序列数据转换为矢量数据。运用训练数据构建基于最小二乘支持向量机的回归估计模型,通过计算实际值与预测值之间的残差来判别异常数据。以成都市一环路北二段进行实证研究,并与传统检测方法比较分析,结果证实该检测方法的有效性。     

Artistic features extraction from chinese calligraphy works via regional guided filter with reference image

Chinese calligraphy is a unique visual art, and and is one of the material basis of China’s traditional cultural heritage. However, time had caused the old calligraphy works to weathering and damages, so it is necessary to utilize advanced technologies to protect those works. One of those technologies is digital imaging, and the obtained images by digital imaging can preserve the visual information of calligraphy works better, furthermore, they can be used in further researches. While the basic works for those researches are to extract the artistic features which include two elements, i.e., form and spirit. However, most of the existing methods only extract the form and ignore the characters’ spirit, especially they are insensitive to the slight variation in complex ink strokes. To solve these problems, this paper proposes an extraction method based on regional guided flter (RGF) with reference images, which is generated by KNN matting and used as the input image for RGF. Since RGF is sensitive to the slight variation of ink, so the detailed information of the inside of strokes can be detected better. Besides, unlike the past works, which filter the whole strokes, RGF filters the inside of strokes and edges in different windows respectively, which results in that the edges are preserved accurately. Results from a deployment of several famous Chinese calligraphy works demonstrate that our method can extract more accurate and complete form and spirit with lower error rate.     

支持力反馈的肝脏手术虚拟仿真及研究

该文提出了一种基于支持力反馈的肝脏虚拟手术仿真系统,支持力反馈的肝脏虚拟手术仿真实现流程主要包括CT影像的肝脏三维重建、空间场景构建、实时动态交互设计、反馈力的手术操作仿真.虚拟仿真环境搭建完成后,用户能够通过头盔式显示设备HTC Vive Pro及其所配备力反馈手柄在虚拟手术场景中进行肝脏手术切割的沉浸式知识学习及实...     

一种基于嵌入式平台的原位测试设备研制

针对现有发动机状态检测设备体积大、重量大,集成度低,导致机动性较差的实际,提出了一种基于嵌入式平台的原位测试设备的设计方案,构建了硬件、软件平台,给出了相关的硬件电路和软件设计.实际应用表明,新研制设备不仅功能齐全,工作稳定可靠,便于地勤人员的维护保障,而且具有使用、携带方便、智能化、集成度高的特点.     

QoS-aware placement of stream processing service

Many emerging online stream processing services require the consideration of quality of service (QoS), which is highly dependent on the placement of services at various hosts. This paper investigates the QoS-aware placement problems of stream processing services under different contexts. On condition that the client demands are stable, the QoS-aware placement problem aiming to minimize the cost when servers are CPU-uncapacitated, is equivalent to the set cover problem, and can be solved by a greedy algorithm with approximation factor O(log?n), where n is the number of clients. However, when CPU capacity constraints on servers are taken into account, the QoS-aware placement problem cannot be approximated unless P=NP. Therefore, we propose two heuristic algorithms: (1) ISCA (Iterated Set Cover-based Algorithm) and (2) KBA (Knapsack-Based Algorithm). We also consider the placement problem of client demands increasing over time. Two objectives, called extension factor and system lifetime, are proposed for demand increment-blind and increment-aware models respectively. Both of them can be solved by extending ISCA and KBA. The experimental results show that ISCA and KBA have distinct effects on different demand sizes. ISCA is more efficient when client demands are relatively small, while KBA performs better for larger demands.     

Subharmonic Resonance Behavior for the Classical Hydrogen Atomic System

Previously unexplored resonance conditions are shown to exist for the classical hydrogen atomic system, where the electron is treated as a classical charged point particle following the nonrelativistic Lorentz-Dirac equation of motion about a stationary nucleus of opposite charge. For circularly polarized (CP) light directed normal to the orbit, very pronounced subharmonic resonance behavior is shown to occur with a variety of interesting properties. In particular, only if the amplitude of the CP light exceeds a critical value, will the resonance continue without radius and energy decay. A perturbation analysis is carried out to illustrate the main features of the behavior. The present phenomena adds to a growing list of other nonlinear dynamical behaviors of this simple system, that may well be important for more deeply understanding classical and quantum connections.     

计算机专业电路课程教学改革与实践

本文针对计算机专业电路课程教学工作中存在的问题,从课程体系、课程内容的调整与选择,教学方法、课程考核方法的改革等方面提出了若干改进措施和建议。     

Is my model any good: differentially private regression diagnostics

Linear and logistic regression are popular statistical techniques for analyzing multi-variate data. Typically, analysts do not simply posit a particular form of the regression model, estimate its parameters, and use the results for inference or prediction. Instead, they first use a variety of diagnostic techniques to assess how well the model fits the relationships in the data and how well it can be expected to predict outcomes for out-of-sample records, revising the model as necessary to improve fit and predictive power. In this article, we develop \(\epsilon \)-differentially private diagnostics tools for regression, beginning to fill a gap in privacy-preserving data analysis. Specifically, we create differentially private versions of residual plots for linear regression and of receiver operating characteristic (ROC) curves as well as binned residual plot for logistic regression. The residual plot and binned residual plot help determine whether or not the data satisfy the assumptions underlying the regression model, and the ROC curve is used to assess the predictive power of the logistic regression model. These diagnostics improve the usefulness of algorithms for computing differentially private regression output, which alone does not allow analysts to assess the quality of the posited model. Our empirical studies show that these algorithms can be effective tools for allowing users to evaluate the quality of their models.