講演抄録/キーワード |
講演名 |
2020-03-16 16:30
ラベル情報を考慮した高速な近似最近傍探索 ○Onur Guler・Yusuke Matsui・Yoichi Sato(東大) PRMU2019-85 |
抄録 |
(和) |
Image retrieval systems commonly assume that the query images have the same distribution with the searched database. However, user queries are often limited by their surroundings and everyday life of the users. For example, a chef may search for kitchen equipment whereas a researcher would search for objects such as a computer and lab equipment. By reducing the search space for the image retrieval task according to user preferences, we can obtain a personalized, faster solution. |
(英) |
Image retrieval systems commonly assume that the query images have the same distribution with the searched database. However, user queries are often limited by their surroundings and everyday life of the users. For example, a chef may search for kitchen equipment whereas a researcher would search for objects such as a computer and lab equipment. By reducing the search space for the image retrieval task according to user preferences, we can obtain a personalized, faster solution. |
キーワード |
(和) |
Image retrieval / User requirements / Personalization / Reduced search space / / / / |
(英) |
Image retrieval / User requirements / Personalization / Reduced search space / / / / |
文献情報 |
信学技報, vol. 119, no. 481, PRMU2019-85, pp. 101-106, 2020年3月. |
資料番号 |
PRMU2019-85 |
発行日 |
2020-03-09 (PRMU) |
ISSN |
Online edition: ISSN 2432-6380 |
著作権に ついて |
技術研究報告に掲載された論文の著作権は電子情報通信学会に帰属します.(許諾番号:10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
PDFダウンロード |
PRMU2019-85 |