PSIDP: Unsupervised deep hashing with pretrained semantic information distillation and preservation

Shi，Yufeng1; You，Xinge1,2; Zhao，Yue3; Xu，Jiamiao4; Ou，Weihua5; Zheng，Feng6; Peng，Qinmu1,2

2022-09-01

10.1016/j.neucom.2022.06.060

NEUROCOMPUTING   Impact Factor & Quartile

0925-2312

1872-8286

Blending free annotation cost, low memory usage and high query speed into a unity, unsupervised deep hashing has shown extraordinary talents in image retrieval. In order to address the lack of semantic information in unsupervised scenario, related works leverage models pretrained on large-scale datasets (e.g., ImageNet) to the estimate semantic similarity for specific datasets (e.g., FLICKR-25 K). However, most of them discard the original semantic attribute in the pretrained datasets, which results in hash codes being over-fitted to the estimated semantic similarity information. To alleviate this problem, this paper tries to consider the semantic similarity of the specific dataset and preserve the semantic attribute from the pretrained dataset simultaneously. Meanwhile, to build a reliable semantic similarity matrix for the specific dataset, this paper develops a similarity distiller by jointly measuring the semantic similarity with the distance between deep features, the distance on the local manifold and the distance on the connectivity graph. Moreover, an efficient attribute preserver is also designed to maintain the correspondence between the hash codes and the category attributes from the pretrained dataset based on regeneration criteria. The method proposed in this paper is named PSIDP for simplicity and extensive image retrieval experiments on four benchmark datasets demonstrate the superiority of the proposed PSIDP method when compared with other state-of-the-art unsupervised deep hashing methods. The code is available at https://github.com/reresearcher/PSIDP.

Content-based image retrieval Pretrained semantic information Unsupervised deep hashing

EI ; SCI

English

Others

National Natural Science Foundation of China["61772220","62101179"] ; Science, Technology and Innovation Commission of Shenzhen Municipality["JCYJ20180305180637611","JCYJ20180305180804836","JSGG20180507182030600"] ; Project of Hubei University School[202011903000002]

Computer Science

Computer Science, Artificial Intelligence

WOS:000931742000001

ELSEVIER

20222712318509

Codes (symbols) ; Content based retrieval ; Hash functions ; Large dataset ; Semantics

Data Processing and Image Processing:723.2 ; Chemical Operations:802.3

COMPUTER SCIENCE

2-s2.0-85133298511

Scopus

1.School of Electronic Information and Communications,Huazhong University of Science and Technology,Wuhan,430074,China
2.Research Institute of Huazhong University of Science and Technology in Shenzhen,Shenzhen,518000,China
3.School of Computer Science and Information Engineering,Hubei University,Wuhan,430062,China
4.Department of Deep Learning,Deeproute Co. Ltd.,Shenzhen,518000,China
5.School of Big Data and Computer Science,Guizhou Normal University,Guiyang,550025,China
6.Department of Computer Science and Engineering,Southern University of Science and Technology,Shenzhen,518055,China

Shi，Yufeng,You，Xinge,Zhao，Yue,et al. PSIDP: Unsupervised deep hashing with pretrained semantic information distillation and preservation[J]. NEUROCOMPUTING,2022,502:1-14.

Shi，Yufeng.,You，Xinge.,Zhao，Yue.,Xu，Jiamiao.,Ou，Weihua.,...&Peng，Qinmu.(2022).PSIDP: Unsupervised deep hashing with pretrained semantic information distillation and preservation.NEUROCOMPUTING,502,1-14.

Shi，Yufeng,et al."PSIDP: Unsupervised deep hashing with pretrained semantic information distillation and preservation".NEUROCOMPUTING 502(2022):1-14.