Learning Decorrelated Hashing Codes With Label Relaxation for Multimodal Retrieval

Due to the correlation among hashing bits, the retrieval performance improvement becomes slower when the hashing code length becomes longer. Existing methods try to regularize the projection matrix as an orthogonal matrix to decorrelate hashing codes. However, the binarization of projected data may...

Full description

Bibliographic Details
Main Authors:	Dayong Tian, Yiwen Wei, Deyun Zhou
Format:	Article
Language:	English
Published:	IEEE 2020-01-01
Series:	IEEE Access
Subjects:	Multimodality hashing binary embedding minimum correlation regularization
Online Access:	https://ieeexplore.ieee.org/document/9072435/

id	doaj-b463594ec597480195859ab64cc6283e
record_format	Article
spelling	doaj-b463594ec597480195859ab64cc6283e2021-03-30T01:35:15ZengIEEEIEEE Access2169-35362020-01-018792607927210.1109/ACCESS.2020.29889239072435Learning Decorrelated Hashing Codes With Label Relaxation for Multimodal RetrievalDayong Tian0https://orcid.org/0000-0002-5830-0650Yiwen Wei1https://orcid.org/0000-0002-8216-6156Deyun Zhou2https://orcid.org/0000-0002-7400-5387School of Electronics and Information, Northwestern Polytechnical University, Xi’an, ChinaSchool of Physics and Optoelectronic Engineering, Xidian University, Xi’an, ChinaSchool of Electronics and Information, Northwestern Polytechnical University, Xi’an, ChinaDue to the correlation among hashing bits, the retrieval performance improvement becomes slower when the hashing code length becomes longer. Existing methods try to regularize the projection matrix as an orthogonal matrix to decorrelate hashing codes. However, the binarization of projected data may completely break the orthogonality. In this paper, we propose a minimum correlation regularization (MCR) for multimodal hashing. Rather than being imposed on projection matrix, MCR is imposed on a differentiable function which approximates the binarization. On the other hand, binary labels could not precisely reflect the distances among data. Hence, we propose a label relaxation scheme to achieve better performance.https://ieeexplore.ieee.org/document/9072435/Multimodalityhashingbinary embeddingminimum correlation regularization
collection	DOAJ
language	English
format	Article
sources	DOAJ
author	Dayong Tian Yiwen Wei Deyun Zhou
spellingShingle	Dayong Tian Yiwen Wei Deyun Zhou Learning Decorrelated Hashing Codes With Label Relaxation for Multimodal Retrieval IEEE Access Multimodality hashing binary embedding minimum correlation regularization
author_facet	Dayong Tian Yiwen Wei Deyun Zhou
author_sort	Dayong Tian
title	Learning Decorrelated Hashing Codes With Label Relaxation for Multimodal Retrieval
title_short	Learning Decorrelated Hashing Codes With Label Relaxation for Multimodal Retrieval
title_full	Learning Decorrelated Hashing Codes With Label Relaxation for Multimodal Retrieval
title_fullStr	Learning Decorrelated Hashing Codes With Label Relaxation for Multimodal Retrieval
title_full_unstemmed	Learning Decorrelated Hashing Codes With Label Relaxation for Multimodal Retrieval
title_sort	learning decorrelated hashing codes with label relaxation for multimodal retrieval
publisher	IEEE
series	IEEE Access
issn	2169-3536
publishDate	2020-01-01
description	Due to the correlation among hashing bits, the retrieval performance improvement becomes slower when the hashing code length becomes longer. Existing methods try to regularize the projection matrix as an orthogonal matrix to decorrelate hashing codes. However, the binarization of projected data may completely break the orthogonality. In this paper, we propose a minimum correlation regularization (MCR) for multimodal hashing. Rather than being imposed on projection matrix, MCR is imposed on a differentiable function which approximates the binarization. On the other hand, binary labels could not precisely reflect the distances among data. Hence, we propose a label relaxation scheme to achieve better performance.
topic	Multimodality hashing binary embedding minimum correlation regularization
url	https://ieeexplore.ieee.org/document/9072435/
work_keys_str_mv	AT dayongtian learningdecorrelatedhashingcodeswithlabelrelaxationformultimodalretrieval AT yiwenwei learningdecorrelatedhashingcodeswithlabelrelaxationformultimodalretrieval AT deyunzhou learningdecorrelatedhashingcodeswithlabelrelaxationformultimodalretrieval
_version_	1724186822136823808

Learning Decorrelated Hashing Codes With Label Relaxation for Multimodal Retrieval

Similar Items