US-Singapore Workshop: Collaborative Research: Understand the World by Analyzing Many Video Streams

NSF CNICUS-新加坡研讨会：合作研究：通过分析许多视频流了解世界许多成千上万的网络摄像机提供近乎实时的世界视图。通过对大量数据的分析，可以提取出有价值的科学信息。然而，大多数数据从未被分析或存档。分析程序必须能够从分布式来源检索数据，管理网络访问，调度计算资源，分配存储空间，分析数据，存储结果，并以易于理解的格式显示结果。虽然数据是可用的，但由于缺乏编写和执行计算机程序的软件框架，目前的分析有限。为了从数据中获取知识，来自六所大学的研究人员正在创建一个可以分析海量视觉数据的系统。该小组将于2014年12月在新加坡组织一次研讨会。在这次研讨会上，团队将分享目前在六所大学独立进行的项目的信息。研究人员将展示和展示他们的项目，讨论和制定合作计划。该团队包括卢永祥(普渡大学负责人)、Ketan Mayer-Patel(北卡罗来纳大学)、刘聪(德克萨斯大学)、魏曾华(新加坡国立大学)、Jason Gu(新加坡技术与设计大学)、温永刚(南洋理工大学)。

NSF CNICUS-Singapore Workshop: Collaborative Research: Understand the World by Analyzing Many Video StreamsMany thousands of network cameras provide near real-time view of the world. Valuable scientific information could be extracted by analyzing the large amounts of data. However, most of the data are never analyzed or archived. Analysis programs must be able to retrieve data from distributed sources, manage network accesses, schedule computation resources, allocate storage space, analyze the data, store the results, and present the results in formats that are easy to comprehend. Although the data are available, analyses today are limited due to the lack of a software framework to write and execute computer programs.To harvest the knowledge from the data, the researchers from six universities are creating a system that can analyze the vast amount of visual data. The team will organize a workshop in December 2014 in Singapore. In this workshop, the team will share the information of the projects currently been conducted independently in the six universities. The researchers will present and demonstrate their projects, discuss and develop a plan for collaboration. The team includes Yung-Hsiang Lu (lead, Purdue University), Ketan Mayer-Patel (University of North Carolina), Cong Liu (University of Texas Dallas), Wei Tsang Ooi (National University of Singapore), Jason Gu (Singapore University of Technology and Design), Yonggang Wen (Nanyang Technological University).