Privacy-preserving image multi-classification deep learning model in robot system of industrial IoT
- PDF / 1,974,926 Bytes
- 18 Pages / 595.276 x 790.866 pts Page_size
- 69 Downloads / 184 Views
(0123456789().,-volV)(0123456789(). ,- volV)
S.I. : HIGHER LEVEL ARTIFICIAL NEURAL NETWORK BASED INTELLIGENT SYSTEMS
Privacy-preserving image multi-classification deep learning model in robot system of industrial IoT Yange Chen1,2 • Yuan Ping2 • Zhili Zhang2
•
Baocang Wang1,2,3 • SuYu He1,3
Received: 22 July 2020 / Accepted: 5 October 2020 Ó Springer-Verlag London Ltd., part of Springer Nature 2020
Abstract Deep learning in robot systems is a popular application that can learn and train the results per requirements, but that collects sensitive information in the training process, easily causing leakage of users’ private information. To date, privacypreserving deep learning models in robot systems have been sparsely researched. To solve the privacy leakage problem of deep learning in robot systems and fill the gap in robotics deep learning privacy research, in this paper a novel privacypreserving image multi-classification deep-learning (PIDL) model in robot systems is presented. In PIDL, two schemes are proposed that adopt two groups of encrypted activation and cost functions—sigmoid plus cross-entropy function (PIDLSC) and softmax plus log-likelihood function (PIDLSL)—with secure calculation protocols, which are applied in a fog control center (FCC) with a non-colluding honest server by homomorphic encryption to improve the training efficiency, solve the encryption computation questions, and protect data and model privacy in robot systems. Security analysis and performance evaluation demonstrate that the proposed schemes realize security, correctness, and efficiency with low communication and computational costs. Keywords Privacy-preserving Deep learning Industrial Internet of Things (IIoT) Secure calculation
1 Introduction Significant achievement and wide application in deep learning have had an important impact in various research areas, especially in Industrial Internet of things (IIoT) application [1, 2], for example, recognition system [3, 4], voice assistants [5], text analysis, image classification [6], language translation. With the appearance of Industry 4.0 and Industry 5.0, deep learning in IIoT plays an important role and has attracted significant attention [7], which has been widely applied in smart power grids [8], robot system, & Zhili Zhang [email protected] 1
State Key Laboratory of Integrated Service Networks, Xidian University, Xi’an, China
2
School of Information Engineering, Xuchang University, Xuchang, China
3
Cryptographic Research Center, Xidian University, Xi’an, China
intelligent control, intelligent vehicle, and intelligent manufacturing. One of the typical IIoT application fields is the robot system. With the emergence of intelligent devices and new robot techniques, current robot systems now face numerous new vulnerabilities, such as new security and privacy issues brought about by deep learning in industrial informatics [9, 10]. Regarding system security, many scholars have explored the cyber security of the robot system. For instance, Dieber et al. [11] enumerated the
Data Loading...
