Video Semantic Models and Their Evaluation Criteria

WANG Yu,ZHOU Li-Zhu,XING Chun-Xiao

DOI: https://doi.org/10.3321/j.issn:0254-4164.2007.03.001

2007-01-01

Chinese Journal of Computers

Abstract:The development of video technology and video-related applications demands strong support in semantic data models.To meet such a requirement,many video semantic data models have been proposed.The semantic model plays a key role in providing query capability and other features for a video database.However,to our knowledge,the criteria for a good semantic model remain open at present.As a result,people lack the rules for evaluating an existing model and the guidelines for the design of a new data model when necessary.To address this issue,this paper proposes twenty two properties as the criteria for video semantic models,and gives the evaluation result of eleven existing rich semantic models according to these properties.It shows that these models mostly concentrate on basic expressive power and query capability,and fulfill users′ primary requirements.But in some advanced features such as expressive power,acquisition and analysis of semantic information,and query capability etc,there are rooms for further enhancement.The paper concludes by indicating some research directions for video semantic models.