Abstract
Due to the risk of data leakage while training deep learning models in a shared environment, we propose a new privacy-preserving deep learning (PPDL) method using a structural image de-identification approach for object classification. The proposed structural image de-identification approach is designed based on the fact that the degree of structural distortion of an image object has the greatest impact on human’s perceptual system. Thus, by modifying only the structural parts of the original one using order preserving encryption(OPE), the proposed structural image de-identification approach decreases only the recognition rate by human. From the experimental results using different standard datasets, we show that the object classification accuracy of the proposed structural image de-identification method is almost the same as the deep learning performance for non-encrypted images, without revealing the original image contents including sensitive information. Also, by handling the trade-off between object classification accuracy and privacy protection for the de-identified image, we experimentally find the optimal size of input image for the proposed structural image de-identification approach.
Highlights
The performance of deep learning has become to exceed human ability in various application services such as language translation service, image recognition, self-driving car service and so on [1], [2]
We find the optimal size of input image in the context of the trade-off between object classification accuracy and privacy; (3) From the evaluation results under various parameters using different well-known standard datasets, we show the effectiveness of the proposed structural image de-identification method
As a set of input images, we used CIFAR-10 [18], which is a standard dataset to evaluate the performance of the deep learning approaches
Summary
The performance of deep learning has become to exceed human ability in various application services such as language translation service, image recognition, self-driving car service and so on [1], [2]. The proposed approach is designed to enable the use of state-of-the-art deep learning techniques for better performance, such as various DNN models and data augmentation, as well as to analyze grayscale images directly for general purposes. The primary objective of the proposed structural image de-identification approach is to protect the privacy of the input data to DNN models for object classification while keeping the high accuracy; (2) We measure trade-off values between utility and privacy according to various parameter values of input image size. We find the optimal size of input image in the context of the trade-off between object classification accuracy and privacy; (3) From the evaluation results under various parameters using different well-known standard datasets, we show the effectiveness of the proposed structural image de-identification method.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
