针对软件工程专业传统实践课程体系中实践能力产出导向不强、实践需求难以满足、评价体系不完善的问题,提出通过更加深入、全面、稳定的校企合作,优化培养模式和重构课程体系.通过湘潭大学软件工程专业教学实践说明该体系教学效果.

持续改进是工程教育专业认证通用标准中核心指标之一,只有贯彻并落实了持续改进的理念,才能真正做到以产出为导向进行培养方案设计,以学生为中心进行课程设计.本研究以湘潭大学通信工程专业为例,通过2017年专业认证入校考察和2020年中期检查的两次实践经历,从基于"胜任力"模型的人才培养模式、一流课程建设和强化专业主体的质量评价监控体系等几个方面,对新形势下通信工程专业的持续改进机制与成效做了研究.

SiamRPN算法采用Ln范数损失训练边界框预测,未考虑预测框与真值框间交并比(inersection over union,IoU)的关系,导致准确性不足.针对该问题,提出一种结合IoU损失的SiamRPN目标跟踪改进算法.设计了 IoU-smooth L1范数联合优化模块,对候选正样本进行IoU损失与smooth L1范数损失的联合优化;依据回归预测结果,用预测框与真值框的IoU作为权重对正样本进行加权分类预测,增加正样本间的区分度,同时确保分类预测与回归预测的关联性.对比实验结果表明:本文所提改进算法能有效提升跟踪性能.

针对目前我国智能驾驶辅助系统识别道路交通标志检测速度慢、识别精度低等问题,提出一种基于YOLOv3的改进的道路交通标志检测算法.首先,将MobileNetv2作为基础特征提取网络引入YOLOv3以形成目标检测网络模块MN-YOLOv3,在MN-YOLOv3主干网络中引入两条Down-up连接进行特征融合,从而减少检测算法的模型参数,提高了检测模块的运行速度,增强了多尺度特征图之间的信息融合;然后,根据交通标志目标形状的特点,使用K-Means++算法产生先验框的初始聚类中心,并在边界框回归中引入距离交并比(DIOU)损失函数来将DIOU与非极大值抑制(NMS)结合;最后,将感兴趣区域(ROI)与上下文信息通过ROI Align统一尺寸后融合,从而增强目标特征表达.实验结果表明,所提算法性能更好,在长沙理工大学中国交通标志检测(CCTSDB)数据集上的平均准确率均值(mAP)可达96.20％.相较于Faster R-CNN、YOLOv3、Cascaded R-CNN检测算法,所提算法拥有具有更好的实时性和更高的检测精度,对各种环境变化具有更好的鲁棒性.

A technique of multiple fault localization based on conditioned execution slicing spectrum considers program dependency and reduces the problem which the efficiency of the technology will decrease with the increase of the number of faults in the program. But the technology is affected by the fault-independent statement and is validated in a particular program. Therefore, Conditioned Classification Execution Slicing Spectrum-based Software Fault Localization(CCESS-SFL) is proposed, The technology improves and classifies predicate conditions in CESS-MFL technology. Firstly, the conditional feature set is determined according to the predicate conditions and the fault-related conditioned execution slicing, the conditional classification is performed according to the conditional feature set to obtain the conditional classification execution slicing spectrum, then, the suspiciousness of the element is calculated, finally the suspiciousness report is generated. CCESS-SFL technology has been validated in seven Siemens suites. It is superior to the current popular Tarantula, Jaccard, Ochiai and CESS-MFL technology and can reduce the impact of fault-independent statements.

在全媒体时代实体书店应采用多种媒体工具自我营销,微信公众号以其低廉的成本和便捷的交互特点,成为实体书店的重要营销工具之一.本文选取40家实体书店为样本,调研其微信公众号的内容建设与运营现状,并提出及时认证、 明确定位、 优化内容、 合理管理、 加强互动等改进策略.

以工程教育专业认证为背景,在结合课堂和实验教学实践的基础上,从工程教育投入和产出的角度出发,本文对《人工智能》课程的教学改革进行了探讨,从教学改革意义、目的、教学内容、教学方法、教学手段等方面进行了具体分析,希望能为同行提供参考.

用户粘性是评价社交社区流量变现能力的重要指标,如何提高用户粘性和忠诚度,从而在逐渐同质化的社交化阅读社区获得竞争优势,已成为各大社交化阅读社区服务商关注的焦点.本文以双系统认知模型对社交化阅读社区用户粘性影响因素进行分析,以微信读书、豆瓣读书、网易云阅读和Zaker为例进行了用户调查,并运用SPSS20.0软件分析了读书社区用户的影响因素,从而提出了激励用户,降低用户的感知成本,优化操作体验等相关针对性措施.

Advertising language is the core of the advertising and a new form of the existent language.There are many kinds of advertising,which include the homophonic words.Homophony is a language phenomenon,which uses the same or similar sound between the words,for example,the phase harmonic relationship,to express the thought.The homophony in Chinese food advertising is extremely rich and has been used frequently.Homophonic words are not only affected by language itself but also physical and psychological factors.

新闻活动中无不渗透着新闻作者的审美情感,然而新闻报道一直讲究追求真实客观。在审美情感渗透的新闻活动中要保持新闻报道的真实性和客观性,必须坚持以事实为基础、以真情动人、理性与情感交融的基本原则。本文试以《人民日报》"走基层"报道为例,分析其新闻作者审美情感的表达技巧。

MCMC-based multi-object visual tracking was investigated here.To improve the confidence of sampling and perform the iteration effectively,a new approach to multi-object visual tracking was proposed based on reversible jump Markov chain Monte Carlo(RJMCMC) sampling.Given image observation,the tracking problem was formulated as computing the MAP(maximum a posteriori) estimation.The prior proposal distribution of object was developed with the aid of association match matrix,and four types of reversible and jump moves were designed for Markov chains dynamics.The likelihood distribution measure was presented via position-weighted colour hist match between reference objects and candidate objects.The state updating was generated from mean-shift(MS) iteration,rather than from random walk in the MCMC sampling.Experimental results and quantitative evaluation demonstrate that the proposed approach is effective for challenge situations.

智慧校园是继数字校园之后关于高校信息化建设的全新概念,也是现代高校管理和发展的必然趋势。但当前,信息分散、共享程度不高仍然是国内许多高校信息化建设普遍存在的问题。本文结合湘潭大学智慧校园建设的实际经验,以湘潭大学自主研发高校教务管理系统为例,详细阐述了数据共享中心的建设以及信息集中与共享方案的具体实施。

陌生化效果的手法是多样的,而新闻叙事视角与聚焦影响着新闻框架的构建、叙事声音、叙事语法、叙事话语、叙事接受等,是陌生化效果产生的首要源泉。本文从新闻叙事视角和聚焦出发,结合具体事例分析新闻媒体在新闻事件基本相同的情况下,采用不同新闻叙事视角和聚焦产生的不同陌生化效果。

This paper proposed an algorithm for pose estimation with an un-calibrated camera based on vector operation.Firstly,it researched the P5P problem,formed five control points into four vectors.According to the process of image formation,the constraint equations in camera pose and intrinsic parameter were set up linearly through vector operation of control points.Then,simplified and determined these equations in terms of linear theory and the orthogonal relation of rotation R.As a result,on the un-calibrated P5P problem,obtained the analytic solutions of camera pose and intrinsic parameter matrix via vector cross or inner product operation.Secondly,studied solution o PnP problem with sufficient constraints.Simulated and real experiments illustrate that the proposed algorithm is effective.

针对经典Mean shift(MS)目标跟踪算法的颜色特征鲁棒差、匹配迭代复杂的缺点,提出一种分层Mean shift (Hierarchical mean shift,HMS)目标跟踪算法.首先通过MS迭代将目标区域特征空间的数据点聚类于模式点,使得以简洁的方式描述前景跟踪目标,建立目标模型与目标候选模型的聚类模式点描述,进行聚类块匹配.然后,导出聚类块模式点匹配下的相似度量函数,进行像素点匹配,结合邻域一致性,计算像素平移量,分层估计序列帧中跟踪目标质心模式点的位置,并给出HMS匹配迭代跟踪算法.实验结果表明,与其他两种MS跟踪算法相比,HMS既能提高序列帧跟踪目标表达与匹配的鲁棒性,又无需匹配所有数据点,算法简洁且有效可行.

Based on image corner points,a digital image watermark algorithm is proposed in this paper.High frequency information from original image is embedded watermark adaptively,with corner points being embedding position.To detect watermark,watermark geometric distortion can be rectified and from which got position to extract watermark accurately according to extracted corner points of marked image.Then number of the effective corner points is statistified to quantitively show the result of performance analysis.After watermark synchrony,correlation is used as judge criterion.Simulation result shows that the method is practicable with higher robustness as shearing and geometric attacks.

This paper proposed a technique for self-calibration based on extended camera model.Changed a conception of thinking from different views photography to different direction perspective projection.Thus,it could obtian three different homography between space plane and image plane from one image simultaneously,and set up the absolute conic equation of circular points within their plane.So a single image completed the process of calibration.Took regular solid to vertify the idea and thinking way.Experiments with simulated data as well as with real images show that the proposed method is workable,and applicable in real applications.

本文设计了一种基于奇异值分解(SVD)的盲检测数字图像水印算法,将原图像分块进行SVD分解,再对每块最大奇异值适当量化并嵌入水印;水印检测时,直接对含水印图像的分块最大奇异值的量化结果的奇偶性判别来提取水印.进一步,通过对图像的不可见性及鲁棒性分析,给出量化强度的理论分析和理论取值.实验表明,该方法可行,算法简洁,抗压缩、滤波、几何攻击性强.

A new model, role access control model based on resource abstract(RD_RBAC), is proposed in the paper. The model is formed from the second abstract of the resource during the course of role authorization and the implementation of security policy. Thus , the reduction of role redundant role and management complexity is obtained. The scheme is vertified by the case in management system of university. Compared with traditional RBAC, RD_RBAC is of more applicability and more security.

Based on Auto-Correlation Function(ACF),a digital image watermark algorithm in the wavelet domain detecting blindly is proposed.High frequency detail information from original image wavelet decomposing is enhanced correlation,then acts as the mask of watermark information.The marked image is obtained by composing the processed content with approximation of the original one.To detect watermark,watermark geometric distortion can be estimated and rectified according to ACF peaks,which is from analyzing detail portion in the marked image wavelet decomposing.After watermark synchrony,correlation is used as judge criterion.Simulation result shows that the method is practicable with higher such robustness as shearing and geometric attacks.