We participated in the user-centered search task at ImageCLEF 2004. In this paper, we proposed two interactive Cross-Language image retrieval systems-T-ICLEF and VCT-ICLEF. The first one is implemented with a practical relevance feedback approach based on textual information while the second one combines textual and image information to help users find a target image. The experimental results show that VCT-ICLEF has a better performance than T-ICLEF in almost all cases. Overall, VCT-ICLEF helps users find the image within a fewer iterations with a maximum of 2 iterations saved.
|期刊||CEUR Workshop Proceedings|
|出版狀態||已發佈 - 2004|
|事件||2004 Cross Language Evaluation Forum Workshop, CLEF 2004, co-located with the 8th European Conference on Digital Libraries, ECDL 2004 - Bath, 英国|
持續時間: 2004 9月 15 → 2004 9月 17
ASJC Scopus subject areas