作者: Christopher Jaynes , Quanren Xiong , R. Matt Steele , Stephen Webb
DOI:
关键词: Complement (set theory) 、 Human–computer interaction 、 Construct (python library) 、 Ground truth 、 Graphical user interface 、 Computer science 、 Component (UML) 、 Task (project management) 、 Application programming interface 、 Dialog box
摘要: We introduce the publicly available Open Development for Video Surveillance (ODViS) system that is specifically designed to support ongoing research in tracking and video surveillance. The provides researchers with ability prototype event recognition techniques, construct manage ground truth data sets, evaluate performance of these systems within a graphical user interface. Passive surveillance are receiving an increasing amount attention both academic industrial laboratories. Meaningful evaluation important but often difficult task. Recent introduction standard datasets systems, corresponding dialog related empirical analysis have made significant progress [1]. ODViS complement efforts by providing Application Programming Interface (API) allows link specific modules into framework, interface visualizing behavior algorithms, component supports comparison different one another or predefined truth.