版权说明 帮助中心
首页 > 成果 > 详情

Sensitivity based image filtering for multi-hashing in large scale image retrieval problems

SCI-EEI
WOS被引频次:2
认领
导出
Link by Springer Journal
反馈
分享
QQ微信 微博
成果类型:
期刊论文
作者:
Ng, Wing W. Y.;Li, Jinchen;Feng, Shaoyong;Yeung, Daniel S.;Chan, Patrick P. K.
通讯作者:
Li, JC
作者机构:
[Ng, Wing W. Y.; Yeung, Daniel S.; Feng, Shaoyong; Chan, Patrick P. K.] School of Computing Science and Engineering, South China University of Technology, Guangzhou, China
[Li, Jinchen] College of Medical Information Engineering, Guangdong Pharmaceutical University, Guangzhou, China
通讯机构:
[Li, Jinchen] Guangdong Pharmaceut Univ, Coll Med Informat Engn, Guangzhou, Guangdong, Peoples R China.
语种:
英文
关键词:
Multi-Hashing;RBFNN;Image filtering;CBIR
期刊:
International Journal of Machine Learning and Cybernetics
ISSN:
1868-8071
年:
2015
卷:
6
期:
5
页码:
777-794
文献类别:
WOS:Article;EI:Journal article (JA)
所属学科:
ESI学科类别:计算机科学;WOS学科类别:Computer Science, Artificial Intelligence
入藏号:
WOS:000361478400006;EI:20153801297669
基金类别:
National Natural Science Foundation of China [61272201]; Program for New Century Excellent Talents in University of China [NCET-11-0162]; Fundamental Research Funds for the Central Universities [2015ZZ023]
机构署名:
本校为通讯机构
院系归属:
医药信息工程学院
摘要:
Hashing is an effective method to retrieve similar images from a large scale database. However, a single hash table requires searching an exponentially increasing number of hash buckets with large Hamming distance for a better recall rate which is time consuming. The union of results from multiple hash tables (multi-hashing) yields a high recall but low precision rate with exact hash code matching. Methods using image filtering to reduce dissimilar images rely on Hamming distance or hash code difference between query and candidate images. However, they treat all hash buckets to be equally important which is generally not true. Different buckets may return different number of images and yield different importance to the hashing results. We propose two descriptors, bucket sensitivity measure and location sensitivity measure, to score both the hash bucket and the candidate images that it contains using a location-based sensitivity measure. A radial basis function neural network (RBFNN) is trained to filter dissimilar images based on the Hamming distance, hash code difference, and the two proposed descriptors. Since the Hamming distance and the hash code difference are readily computed by all hashing-based image retrieval methods, and both the RBFNN and the two proposed sensitivity-based descriptors are computed offline when hash tables become available, the proposed sensitivity based image filtering method is efficient for a large scale image retrieval. Experimental results using four large scale databases show that the proposed method improves precision at the expense of a small drop in the recall rate for both data-dependent and data-independent multi-hashing methods as well as multi-hashing combining both types.
参考文献:
Lv YM, 2015, IEEE T MULTIMEDIA, V17, P1225, DOI 10.1109/TMM.2015.2437712
Li P, 2013, IEEE T MULTIMEDIA, V15, P141, DOI 10.1109/TMM.2012.2199970
Salakhutdinov R, 2009, INT J APPROX REASON, V50, P969, DOI 10.1016/j.ijar.2008.11.006
Li P, 2013, NEUROCOMPUTING, V120, P83, DOI 10.1016/j.neucom.2012.07.053
Bian W, 2010, IEEE T IMAGE PROCESS, V19, P545, DOI 10.1109/TIP.2009.2035223

反馈

验证码:
看不清楚,换一个
确定
取消

成果认领

标题:
用户 作者 通讯作者
请选择
请选择
确定
取消

提示

该栏目需要登录且有访问权限才可以访问

如果您有访问权限,请直接 登录访问

如果您没有访问权限,请联系管理员申请开通

管理员联系邮箱:yun@hnwdkj.com