Detail of Publication
| Text Language | Japanese |
|---|---|
| Authors | Kinji Takeuchi,Koichi Kise |
| Title | Text-Based Image Retrieval Using Earth Mover's Distance |
| Journal | 情報処理学会研究報告 |
| No. | NL-177-5 |
| Reviewed or not | Not reviewed |
| Month & Year | January 2007 |
| Abstract | In this report, we propose a method of Text-Based Image Retrieval(TBIR) using Earth Mover's Distance(EMD). Our approach uses EMD to combine textual and visual features into one semantic space, to improve the performance of TBIR. We compare our method with four conventional methods including LSI and PLSI. The results show our method is superior to others in average precision. We also applied an approximation in computing EMD for improving the efficiency of the proposed method. The results show that the approximate method can reduce the computational complexity in about 8%, with little loss of average precision. |
- Following file is available.
- Entry for BibTeX
@InCollection{Takeuchi2007, author = {Kinji Takeuchi and Koichi Kise}, title = {Text-Based Image Retrieval Using Earth Mover's Distance}, booktitle = {情報処理学会研究報告}, year = 2007, month = jan, number = {NL-177-5} }