Protecting Privacy in Shared Photos via Adversarial Examples Based Stealth

Yujia Liu,Weiming Zhang,Nenghai Yu

doi:10.1155/2017/1897438

Yujia Liu, Weiming Zhang + Show 1 more

Open Access

https://doi.org/10.1155/2017/1897438

Copy DOI

Abstract

Online image sharing in social platforms can lead to undesired privacy disclosure. For example, some enterprises may detect these large volumes of uploaded images to do users’ in-depth preference analysis for commercial purposes. And their technology might be today’s most powerful learning model, deep neural network (DNN). To just elude these automatic DNN detectors without affecting visual quality of human eyes, we design and implement a novel Stealth algorithm, which makes the automatic detector blind to the existence of objects in an image, by crafting a kind of adversarial examples. It is just like all objects disappear after wearing an “invisible cloak” from the view of the detector. Then we evaluate the effectiveness of Stealth algorithm through our newly defined measurement, named privacy insurance. The results indicate that our scheme has considerable success rate to guarantee privacy compared with other methods, such as mosaic, blur, and noise. Better still, Stealth algorithm has the smallest impact on image visual quality. Meanwhile, we set a user adjustable parameter called cloak thickness for regulating the perturbation intensity. Furthermore, we find that the processed images have transferability property; that is, the adversarial images generated for one particular DNN will influence the others as well.

Highlights

With the pervasiveness of cameras, especially smartphone cameras, coupled with the almost ubiquitous availability of Internet connectivity, it is extremely easy for people to capture photos and share them on social networks
Our experiment shows that Stealth algorithm far outdoes several common methods of disturbing image, no matter in effectiveness or in image visual quality
Results indicate that our method works best and has minimal impact on car : 0.995 car : 0.7c5a2r :c0a.r84: 70.97c5ar : 0.961 car : 0.996 person : 0.990person : 0.946 person : 0.931 person : 0.944 person : 0.983 person : 0.976 person : 0.991 dinbiontgtlteab:bl0oe.t9:t0l0e2.9: 909.708 chair : 0.983 diningtable : 0.667 person : 0.997 cow : 0.887 cow : 0.998 bird : 0.824 (a) image visual quality. (iii) We explore the relations among cloak thickness, visual quality, and privacy insurance in the algorithm. (iv) We illustrate the transferability of our Stealth algorithm on different deep neural network (DNN)

Summary

Introduction

With the pervasiveness of cameras, especially smartphone cameras, coupled with the almost ubiquitous availability of Internet connectivity, it is extremely easy for people to capture photos and share them on social networks. A portrait privacy preserving photo capturing and sharing system can give users, who are photographed, the selection to choose appearing (select the “tagged” item) in the photo or not (select the “invisible” item) [14] These processing methods can be good ways to shield people’s access. The traditional processing methods (mosaic, blur, etc.) will greatly reduce image quality undesirably, and not work well to the automatic detection system based on DNN, as shown in the later experimental results (Figure 6). (i) We realize the privacy protection for image content by means of resisting automatic detection machine based on deep neural networks.

Related Work

Object Detectors Based on DNNs

Stealth Algorithm for Privacy

Experiment and Evaluation

Findings

Conclusion and Future Work