Vol.:(0123456789) 1 3 Journal of Ambient Intelligence and Humanized Computing https://doi.org/10.1007/s12652-020-02565-z ORIGINAL RESEARCH Secure image classifcation with deep neural networks for IoT applications Abdelrhman Hassan 1  · Fei Liu 1  · Fanchuan Wang 1  · Yong Wang 1 Received: 27 March 2020 / Accepted: 17 September 2020 © Springer-Verlag GmbH Germany, part of Springer Nature 2020 Abstract The Internet-of-Things (IoT) are used everywhere in our daily lives. IoT applications provide us with many useful functionali- ties such as preventing fres, detecting and tracking objects, controlling and reporting the changes in/outside the environments, and capturing images/videos in our homes, roads, and ofces. For example, the images data gathered through the smart sensors of autonomous vehicles can serve in various applications such as trafc monitoring, prediction of road conditions, and classifcation of objects. Image classifcation with deep neural networks (DNNs) on the cloud is such a machine learn- ing task and has great market potentials for IoT applications. Nevertheless, the deployment of these “smart” IoT devices and applications can raise the risks of security issues. It still sufers from the challenges of relieving IoT devices from excessive computation burdens, such as data encryption, feature extraction, and image classifcation. In this paper, we propose and implement an indistinguishability-chosen plaintext attack secure image classifcation framework with DNN for IoT Applica- tions. The framework performs a secure image classifcation on the cloud without the IoT device’s constant interaction. We propose and implement a real number computation mechanism and a divide-and-conquer mechanism for the secure evaluation of linear functions in DNNs, as well as a set of unifed ideal protocols for the evaluation of non-linear functions in DNNs. The information about the image contents, the private DNNs model parameters and the intermediate results is strictly concealed by the conjunctive use of the lattice-based homomorphic scheme and 2-PC secure computation techniques. A pre-trained deep convolutional neural network model, i.e., Visual Geometry Group (VGG-16), is used to extract the deep features of an image. The comprehensive experimental results show that our framework is efcient and accurate. In addition, we evaluate the security of our framework by performing the white-box membership inference attack which is believed to be the most powerful attack on DNNs models. The failure of the attack indicates that our framework is practical secure. Keywords Internet-of-Things · Image classifcation · Convolutional neural network · Lattice-based homomorphic scheme · Secure multiparty computation 1 Introduction Recent advances in Internet-of-Things (IoT) have resulted in billions of smart devices connecting to the Internet. These devices can frequently produce data from the physical envi- ronment and permit data fow through wireless commu- nication technologies (Sicari et al. 2015). Indeed, several applications such as Smart city (Elhabob et al. 2018), IIoT (Elhabob et al. 2020), and Vehicle ad hoc networks (Toor et al. 2008; Elhabob et al. 2019), have been developed based on the inter-connection of these smart devices. Autonomous/ Self-Driving Vehicles (AVs/SDVs) is a signifcant area that IoT has revolutionized (Khayyam et al. 2020; Joy et al. 2018). The network of the AVs incorporates vehicles that are IoT-enabled via the integration of data within the network. AVs have equipped with smart IoT devices and sensors such as cameras, GPS, and Radars that frequently generated a real-time data from the AVs environments. Such data can be gathered from its sensors shared with another AVs; others can be collected via the IoT platform from various gateways (e.g., parking, trafc information, and environmental infor- mation) (Khayyam et al. 2020). * Yong Wang cla@uestc.edu.cn Abdelrhman Hassan abdhassan25@yahoo.com 1 Center for Cyber Security, School of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu 611731, China