Advertising to the masses: the effects of crowding on the attention to place-based advertising

Advertising situated in environments where people congregate or pass through on their way to somewhere else benefit from being placed in such high traffic areas. However, these strategically placed ads also suffer from conditions of human crowding that prevents them from being noticed and processed. We undertake a study of place-based advertising in a shopping mall using facial recognition software to determine the effect of human density on the attention directed to advertising. We find that as human density increases, attention to advertising decreases, but only to a point where it begins to increase again. Our research also finds that human density plays a moderating role on the motivation to process advertising.     

采用航迹特征的智能机型识别

针对空战目标识别中机型自动识别比较困难的问题，提出了采用航迹特征的智能目标识别方法。利用卷积神经网络（Convolutional Neural Network，CNN）分层学习特征的能力，训练CNN算法模型自动地从航迹数据中学习有用的特征并分类。利用沿海实地采集的15个类别的飞机航迹数据，经一系列数据预处理后作为智能识别算法的训练和测试数据，在验证实验中描述了算法网络的相关配置,对比了CNN与其他分类器的识别结果。实验结果表明，CNN具有很好的识别性能。     

Blockchain for tracking serial numbers in money exchanges

Money exchange is one of the most common day‐to‐day activities performed by humans in the daily market. This paper presents an approach to money tracking through a blockchain. The proposed approach consists of three main components: serial number localization, serial number recognition, and a blockchain to store all transactions and ownership transfers. The approach was tested with a total of 110 banknotes of different currency types and achieved an average accuracy of 91.17%. We conducted a user study in real‐time with 21 users, and the mean accuracy across all users was 86.42%. Each user gave us feedback on the proposed approach, and most of them welcomed the idea.     

一种基于位同步的复杂调制样式识别方法

针对卫星数字化视频广播第二代标准（Digital Video Broadcasting-Satellite -Second Generation，DVB-S2）中的多进制幅度移相键控（Multiple Amplitude and Phase-Shift Keying，MAPSK）和多进制正交幅度调制（Multiple Quadrature Amplitude Modulation，MQAM）信号的调制识别，提出了基于位同步的识别方法。首先利用信号星座图特征对信号包络进行位同步，然后统计信号码元的幅度值个数及其幅度值分布提取统计特征参数，最后利用特征参数完成识别。仿真结果表明，所提算法在载波频率等参数未知的条件下，能够有效对信号进行调制识别。     

Recognition versus disclosure of finance leases: Evidence from Japan

This study examines whether credit market participants—bond investors and credit rating agencies—treat recognized and disclosed finance leases differently when assessing firms’ credit risk in Japan. I use firms’ credit risk, measured by bond spreads and credit ratings, to investigate the relations between recognized versus disclosed finance lease obligations and firms’ credit risk following the adoption of Statement No. 13, Accounting Standard for Lease Transactions. For a sample of firms issuing new bonds, I find that, unlike recognized finance leases, disclosed finance leases are not associated with bond spreads. Moreover, the associations between recognized versus disclosed finance leases and bond spreads are substantially different. Conversely, recognized and disclosed finance leases are associated with credit ratings and are processed similarly when credit ratings are determined. Taken together, my results suggest that the sophistication of capital market participants influences their credit risk assessments of recognized versus disclosed finance leases.     

空间目标的联合矩阵束极点估计

极点作为谐振区雷达目标的最主要电磁特性，具有很大的空间目标识别前景。但是在使用频域观测数据提取目标极点时，存在虚假极点和估计精度问题。针对上述问题，首先通过加窗和截断改进了后时响应信号的获取流程，抑制了虚假极点影响；然后采用最小描述长度法（MDL）估计极点个数，避免了定阶错误引起的虚假极点；最后采用联合矩阵束法估计极点，有效减少了单角度数据估计极点时存在的遗漏极点和虚假极点问题。FEKO软件验证表明，所提算法具有更高的精度和抗噪能力。     

基于LBPH算法的人脸识别算法研究与设计

人脸识别技术是一种根据人脸的特征信息进行身份判定的技术,此技术经过几十年的发展,由弱人工智能向强人工智能转化,论文提出了一种Python语言利用LBPH算法提取人脸特征、实现人脸识别的方法。     

How entrepreneurs recognise entrepreneurial opportunity and its gaps: a cognitive theory perspective

ABSTRACT

Previous studies have verified the relationship between entrepreneurial opportunity recognition and performance. However, the influences of a perceived gap between entrepreneurial opportunities before and after launching a new venture have been neglected. This study applied the partial least square method to analyse the entrepreneurial behaviours and performance of 214 Taiwanese new ventures, and adapted the bootstrap re-sampling method for estimating the variables. The empirical results showed that entrepreneurial resource and entrepreneurial alertness significantly influence the perceived opportunity, while the entrepreneurial resource significantly affects the actual opportunity. The results further demonstrated that both the perceived opportunity and actual opportunity considerably influence new venture performance. This study offered a research framework to better understand the cognition path of entrepreneurial opportunity, and suggested the factors affecting the recognition of perceived opportunity by longitudinal method can be further evaluated. To improve performance, entrepreneurs should enrich entrepreneurial resources and allocate them appropriately, and heighten and maintain their alertness to reduce the perceived gap arising from recognising entrepreneurial opportunities.     

深度学习技术在海关风险甄别中的应用研究

随着互联网的高速发展和人工智能时代的到来,越来越多从前必须由人脑完成的工作能够利用计算机技术来完成,而深度学习的出现更解决了传统机器学习算法在计算机视觉领域、自然语言处理领域表现不佳的问题,使机器也能够拥有准确感知图像和语音的能力。人脸识别是深度学习网络最常见的应用场景之一,具有自然、直接、方便的特点,且不需要检测对象配合,因此非常适合用于公共安全领域的风险检测。研究充分结合海关实际需求,搭建基于深度学习技术的人脸识别模型,提供对通关旅客进行实时风险甄别的解决方案,以及海关通关风险防控场景的理论参考,为后续深度学习技术在海关业务的研究提供支撑。     

动态能力理论视角下战略柔性对企业绩效的影响研究---差异化动态环境规制强度的调节效应

当前企业推进绿色转型进程、提升企业绩效,是实现高质量发展的重要途径。鉴于企业主动应对内外部因素联动影响的重要性,以动态能力理论视角,在探讨了企业变革发展驱动机制的基础上,尝试构建动态环境规制嵌入下的战略柔性与企业绩效的理论框架。分析并实证检验了战略柔性对企业绩效以及差异化强度的动态环境规制在二者关系中调节效应的作用机理和边界。研究结果显示,当前情境下,前瞻柔性、响应柔性对企业绩效均有显著的正向影响,且响应柔性的贡献度较高;两者交互效应亦能显著的正向影响企业绩效;表明企业对环境变化的预测、反应能力以及将两者协同配合是绿色转型发展的重要措施。一般强度的动态环境规制在前瞻柔性、响应柔性与企业绩效的关系中均不起显著的调节效应;高强度动态环境规制仅在响应柔性和企业绩效的关系中起显著负向调节效应;表明动态环境下,一般强度的政府规制政策对企业的决策和创新等行为更多的是无显著影响,充分尊重其自主发展;同时政府要根据环境变化,合理控制环境规制实施强度,谨防给企业应对环境冲击造成额外的负担。综合研究结果,提出企业加强多重柔性机制建设的力度和政府注重环境规制政策引导作用的策略,最终实现各利益相关者协同推进高质量发展的目标。