AUDIO INDEXING AND RETRIEVAL IN MPEG-1 COMPRESSED DOMAIN

摘要：

This paper proposes a novel audio retrieval method using a two-resolution-level index scheme in MPEG-1 compressed domain. A window-resolution-level index extracted in classification step is used to locate the possible matches. A more accurate searching procedure on group resolution level is performed only if there is a ‘possible’ match between query and database. Experimental results show that the proposed method can achieve a better precision and recall rate than traditional audio retrieval method. And dramatically reduce the computational time.

关键词： MPEG-1 audio compressed domain hierarchical retrieval multidimensional index scalefactors.

作者: Miao Li Jiqing. Han

作者单位: School of Computer Science and Technology,Harbin Institute of Technology,Harbin,China

会议类型: 国际会议

会议名称: 2008高等智能国际会议(2008 International Conference on Advanced Intelligence)

会议地点: 北京

会议语种:英文

在线出版日期: 2008-10-18（万方平台首次上网日期，不代表论文的发表时间）