Mobile visual search algorithm based on improved VGG-F and hash with application in IoT. (15th May 2023)

• Record Type: Journal Article
• Authors: Ji, Shanshan; Li, Jianxin; Liu, Jie; Cao, Wenliang; Li, Bin; Jiang, Fei; Liu, Yang
• Is Part Of: International journal of grid and utility computing. Volume 14:Number 2/3(2023)
• Page Start: 94
• Page End: 106
• Publication Date: 2023-05-15
• Subjects: IoT -- deep learning -- hash function -- VGG-F -- mobile visual search -- library digital humanities; Electronic data processing -- Distributed processing -- Periodicals; Electronic commerce -- Management -- Computer programs -- Periodicals; 004.605
• Languages: English
• ISSNs: 1741-847X

Abstract

This paper explores the use of deep learning based hash method in IoT to build a more powerful and real-time mobile visual search, and proposes a lightweight, low latency and high-precision mobile visual search algorithm based on deep hash method and its application in IoT. This paper discusses two aspects: image semantic feature extraction and fast retrieval. Based on the deep learning method and hash method, an image semantic feature extraction model for Library Digital Humanities is constructed, and the loss function applicable to the field is constructed. The model training and MVS retrieval process experiment are carried out through the Library Data set, and this method is applied to the library IoT system. In order to verify the effectiveness of our method, we compared the proposed method with VHB, SSFS, DLBH, SSDH and other methods. Experimental results show that the proposed method is more effective and robust than the existing methods.