Query expansion with pairwise learning in object retrieval challenge

Hao Liu, Atsushi Shimada, Xing Xu, Hajime Nagahara, Hideaki Uchiyama, Rin-Ichiro Taniguchi

研究成果: 著書/レポートタイプへの貢献会議での発言

1 引用 (Scopus)

抄録

Making a reasonable ranking on images in dataset is one of the main objectives for object retrieval challenge, and in this paper we intend to improve the ranking quality. We follow the idea of query expansion in previous researches. Based on the use of bag-of-visual-words model, tf-idf scoring and spatial verification, previous method applied a pointwise style learning in query expansion stage, using but not fully exploring verification results. We intend to extend their learning approach for better discriminative power in retrieval. In re-ranking stage we propose a method using pairwise learning, instead of pointwise learning previously used. We could obtain more reliable ranking on a shortlist of examples. If this verification itself is reliable, a good re-ranking should best preserve this sub-ranking order. Thus in our proposed method, we are motivated to leverage a pairwise learning method to incorporate the ranking sequential information more efficiently. We evaluate and compare our proposed method with previous methods over Oxford 5k dataset, a standard benchmark dataset, where our method achieve better mean average precision and showed better discriminative power.

元の言語英語
ホスト出版物のタイトル2015 Frontiers of Computer Vision, FCV 2015
出版者Institute of Electrical and Electronics Engineers Inc.
ISBN(電子版)9781479917204
DOI
出版物ステータス出版済み - 1 1 2015
イベント2015 21st Korea-Japan Joint Workshop on Frontiers of Computer Vision, FCV 2015 - Mokpo, 大韓民国
継続期間: 1 28 20151 30 2015

その他

その他2015 21st Korea-Japan Joint Workshop on Frontiers of Computer Vision, FCV 2015
大韓民国
Mokpo
期間1/28/151/30/15

All Science Journal Classification (ASJC) codes

  • Computer Vision and Pattern Recognition

これを引用

Liu, H., Shimada, A., Xu, X., Nagahara, H., Uchiyama, H., & Taniguchi, R-I. (2015). Query expansion with pairwise learning in object retrieval challenge. : 2015 Frontiers of Computer Vision, FCV 2015 [7103703] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/FCV.2015.7103703

Query expansion with pairwise learning in object retrieval challenge. / Liu, Hao; Shimada, Atsushi; Xu, Xing; Nagahara, Hajime; Uchiyama, Hideaki; Taniguchi, Rin-Ichiro.

2015 Frontiers of Computer Vision, FCV 2015. Institute of Electrical and Electronics Engineers Inc., 2015. 7103703.

研究成果: 著書/レポートタイプへの貢献会議での発言

Liu, H, Shimada, A, Xu, X, Nagahara, H, Uchiyama, H & Taniguchi, R-I 2015, Query expansion with pairwise learning in object retrieval challenge. : 2015 Frontiers of Computer Vision, FCV 2015., 7103703, Institute of Electrical and Electronics Engineers Inc., 2015 21st Korea-Japan Joint Workshop on Frontiers of Computer Vision, FCV 2015, Mokpo, 大韓民国, 1/28/15. https://doi.org/10.1109/FCV.2015.7103703
Liu H, Shimada A, Xu X, Nagahara H, Uchiyama H, Taniguchi R-I. Query expansion with pairwise learning in object retrieval challenge. : 2015 Frontiers of Computer Vision, FCV 2015. Institute of Electrical and Electronics Engineers Inc. 2015. 7103703 https://doi.org/10.1109/FCV.2015.7103703
Liu, Hao ; Shimada, Atsushi ; Xu, Xing ; Nagahara, Hajime ; Uchiyama, Hideaki ; Taniguchi, Rin-Ichiro. / Query expansion with pairwise learning in object retrieval challenge. 2015 Frontiers of Computer Vision, FCV 2015. Institute of Electrical and Electronics Engineers Inc., 2015.
@inproceedings{e29d3e4f4a644068a2fc7481436efed6,
title = "Query expansion with pairwise learning in object retrieval challenge",
abstract = "Making a reasonable ranking on images in dataset is one of the main objectives for object retrieval challenge, and in this paper we intend to improve the ranking quality. We follow the idea of query expansion in previous researches. Based on the use of bag-of-visual-words model, tf-idf scoring and spatial verification, previous method applied a pointwise style learning in query expansion stage, using but not fully exploring verification results. We intend to extend their learning approach for better discriminative power in retrieval. In re-ranking stage we propose a method using pairwise learning, instead of pointwise learning previously used. We could obtain more reliable ranking on a shortlist of examples. If this verification itself is reliable, a good re-ranking should best preserve this sub-ranking order. Thus in our proposed method, we are motivated to leverage a pairwise learning method to incorporate the ranking sequential information more efficiently. We evaluate and compare our proposed method with previous methods over Oxford 5k dataset, a standard benchmark dataset, where our method achieve better mean average precision and showed better discriminative power.",
author = "Hao Liu and Atsushi Shimada and Xing Xu and Hajime Nagahara and Hideaki Uchiyama and Rin-Ichiro Taniguchi",
year = "2015",
month = "1",
day = "1",
doi = "10.1109/FCV.2015.7103703",
language = "English",
booktitle = "2015 Frontiers of Computer Vision, FCV 2015",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
address = "United States",

}

TY - GEN

T1 - Query expansion with pairwise learning in object retrieval challenge

AU - Liu, Hao

AU - Shimada, Atsushi

AU - Xu, Xing

AU - Nagahara, Hajime

AU - Uchiyama, Hideaki

AU - Taniguchi, Rin-Ichiro

PY - 2015/1/1

Y1 - 2015/1/1

N2 - Making a reasonable ranking on images in dataset is one of the main objectives for object retrieval challenge, and in this paper we intend to improve the ranking quality. We follow the idea of query expansion in previous researches. Based on the use of bag-of-visual-words model, tf-idf scoring and spatial verification, previous method applied a pointwise style learning in query expansion stage, using but not fully exploring verification results. We intend to extend their learning approach for better discriminative power in retrieval. In re-ranking stage we propose a method using pairwise learning, instead of pointwise learning previously used. We could obtain more reliable ranking on a shortlist of examples. If this verification itself is reliable, a good re-ranking should best preserve this sub-ranking order. Thus in our proposed method, we are motivated to leverage a pairwise learning method to incorporate the ranking sequential information more efficiently. We evaluate and compare our proposed method with previous methods over Oxford 5k dataset, a standard benchmark dataset, where our method achieve better mean average precision and showed better discriminative power.

AB - Making a reasonable ranking on images in dataset is one of the main objectives for object retrieval challenge, and in this paper we intend to improve the ranking quality. We follow the idea of query expansion in previous researches. Based on the use of bag-of-visual-words model, tf-idf scoring and spatial verification, previous method applied a pointwise style learning in query expansion stage, using but not fully exploring verification results. We intend to extend their learning approach for better discriminative power in retrieval. In re-ranking stage we propose a method using pairwise learning, instead of pointwise learning previously used. We could obtain more reliable ranking on a shortlist of examples. If this verification itself is reliable, a good re-ranking should best preserve this sub-ranking order. Thus in our proposed method, we are motivated to leverage a pairwise learning method to incorporate the ranking sequential information more efficiently. We evaluate and compare our proposed method with previous methods over Oxford 5k dataset, a standard benchmark dataset, where our method achieve better mean average precision and showed better discriminative power.

UR - http://www.scopus.com/inward/record.url?scp=84937154711&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84937154711&partnerID=8YFLogxK

U2 - 10.1109/FCV.2015.7103703

DO - 10.1109/FCV.2015.7103703

M3 - Conference contribution

AN - SCOPUS:84937154711

BT - 2015 Frontiers of Computer Vision, FCV 2015

PB - Institute of Electrical and Electronics Engineers Inc.

ER -