Deep hash image retrieval based on supervised semantic preservation

Yulong Xia; Hong Yu; Shuyin Xia; Hongyuan Shan; Wanming Yang

doi:10.1117/12.2624163

19 December 2021 Deep hash image retrieval based on supervised semantic preservation

Yulong Xia, Hong Yu, Shuyin Xia, Hongyuan Shan, Wanming Yang

Proceedings Volume 12128, Second International Conference on Industrial IoT, Big Data, and Supply Chain; 121280L (2021) https://doi.org/10.1117/12.2624163
Event: 2nd International Conference on Industrial IoT, Big Data, and Supply Chain, 2021, Macao, China

Abstract

In recent years, with the rapid increase of image data in the Internet, the requirements of image data storage and retrieval are increasingly high. Tra-ditional image retrieval is based on high-dimensional features of the image to carry out similarity (Euclidean distance or cosine distance) calculation for retrieval. Storage and retrieval cost is high, although there are some methods to achieve hash retrieval, but the accuracy is generally not very high. In this paper, a deep hashing coding method based on supervised learning is proposed. This method uses deep neural network to obtain approximate binary codes. After quantifying these approximate binary codes, it can simply and quickly search similar images from massive image data, thus realizing large-scale image retrieval technology. We are in general MNIST dataset, CIFAR-10 dataset, SUN397 dataset and large-scale visual dataset ILSVCR2012 commonly used evaluation test data sets, such as the experimental results showed that the proposed method can achieve good retrieval accuracy, the MAP value of CIFAR-10 compared with existing methods improved a lot, to prove the effectiveness of the proposed method.

Citation Download Citation

Yulong Xia, Hong Yu, Shuyin Xia, Hongyuan Shan, and Wanming Yang "Deep hash image retrieval based on supervised semantic preservation", Proc. SPIE 12128, Second International Conference on Industrial IoT, Big Data, and Supply Chain, 121280L (19 December 2021); https://doi.org/10.1117/12.2624163

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available