An efficient comparison of two indexing-based deep learning models for the formation of a web-application based IoT-cloud network

Abstract

To search a particular image file amongst a large and massive encrypted database is very time consuming and hectic task. Many encryption and searching techniques have been used but they did not prove effective to support smart devices in order to provide input image and retrieve the required results on the personal gadgets of the user. Therefore, based on these facts, an intelligent and advanced multimodal system has been developed in this paper which is based on encrypted content-based search. Thus, in order to perform content-based search two type of novel deep learning techniques, namely cluster-based deep belief network and supervised similarity-based convolutional neural network have been used. The proposed models have been influenced by special indexing techniques to retrieve the best relevant and similar images in very less time. In order to secure the entire images of the database, confusion-diffusion technique based on chaotic map encryption has been used. In order to develop the internet of things model and to support smart device users, a web based application has also been developed using Apache Tomcat server and linking between java and MATLAB has been done using MATLAB engine. Analysis of many parameters like precision, recall, f-score, entropy, correlation coefficient and time has been done here. Also, the proposed system has been compared to many latest and related techniques by using two benchmark and renowned datasets namely WANG and COIL-100.