Video Retrieval Using High Level Features: Exploiting Query Matching and Confidence-Based Weighting

@incollection{citeulike:2709844, abstract = {Recent research in video retrieval has focused on automated, high-level feature indexing on shots or frames. One important application of such indexing is to support precise video retrieval. We report on extensions of this semantic indexing on news video retrieval. First, we utilize extensive query analysis to relate various high-level features and query terms by matching the textual description and context in a time-dependent manner. Second, we introduce a framework to effectively fuse the relation weights with the detectors’ confidence scores. This results in individual high level features that are weighted on a per-query basis. Tests on the TRECVID 2005 dataset show that the above two enhancements yield significant improvement in performance over a corresponding state-of-the-art video retrieval baseline.}, author = {Neo, Shi-Yong and Zhao, Jin and Kan, Min-Yen and Chua, Tat-Seng }, citeulike-article-id = {2709844}, doi = {10.1007/11788034\_15}, journal = {Image and Video Retrieval}, keywords = {acm\_mm, retrieval, video}, pages = {143--152}, posted-at = {2008-04-23 20:36:38}, priority = {2}, title = {Video Retrieval Using High Level Features: Exploiting Query Matching and Confidence-Based Weighting}, url = {http://dx.doi.org/10.1007/11788034\_15}, year = {2006} }

TY - CHAP ID - citeulike:2709844 L3 - citeulike-article-id:2709844 TI - Video Retrieval Using High Level Features: Exploiting Query Matching and Confidence-Based Weighting JF - Image and Video Retrieval SP - 143 EP - 152 N2 - Recent research in video retrieval has focused on automated, high-level feature indexing on shots or frames. One important application of such indexing is to support precise video retrieval. We report on extensions of this semantic indexing on news video retrieval. First, we utilize extensive query analysis to relate various high-level features and query terms by matching the textual description and context in a time-dependent manner. Second, we introduce a framework to effectively fuse the relation weights with the detectors’ confidence scores. This results in individual high level features that are weighted on a per-query basis. Tests on the TRECVID 2005 dataset show that the above two enhancements yield significant improvement in performance over a corresponding state-of-the-art video retrieval baseline. KW - acm_mm KW - retrieval KW - video AU - Neo, Shi-Yong AU - Zhao, Jin AU - Kan, Min-Yen AU - Chua, Tat-Seng PY - 2006/// UR - http://dx.doi.org/10.1007/11788034_15 ER -