Image Privacy Prediction Using Deep Neural Networks

Abstract

Images today are increasingly shared online on social networking sites such as Facebook, Flickr, Foursquare, and Instagram. Image sharing occurs not only within a group of friends but also more and more outside a user's social circles for purposes of social discovery. Despite that current social networking sites allow users to change their privacy preferences, this is often a cumbersome task for the vast majority of users on the Web, who face difficulties in assigning and managing privacy settings. When these privacy settings are used inappropriately, online image sharing can potentially lead to unwanted disclosures and privacy violations. Thus, automatically predicting images' privacy to warn users about private or sensitive content before uploading these images on social networking sites has become a necessity in our current interconnected world.In this paper, we explore learning models to automatically predict appropriate images' privacy as private or public using carefully identified image-specific features. We study deep visual semantic features that are derived from various layers of Convolutional Neural Networks (CNNs) as well as textual features such as user tags and deep tags generated from deep CNNs. Particularly, we extract deep (visual and tag) features from four pre-trained CNN architectures for object recognition, i.e., AlexNet, GoogLeNet, VGG-16, and ResNet, and compare their performance for image privacy prediction. Among all four networks, we observe that ResNet produces the best feature representations for this task. We also fine-tune the pre-trained CNN architectures on our privacy dataset and compare their performance with the models trained on pre-trained features. The results show that even though the overall performance obtained using the fine-tuned networks is comparable to that of pre-trained networks, the fine-tuned networks provide an improved performance for the private class as compared to models trained on the pre-trained features. Results of our experiments on a Flickr dataset of over thirty thousand images show that the learning models trained on features extracted from ResNet outperform the state-of-the-art models for image privacy prediction. We further investigate the combination of user tags and deep tags derived from CNN architectures using two settings: (1) SVM on the bag-of-tags features; and (2) text-based CNN. We compare these models with the models trained on ResNet visual features obtained for privacy prediction. Our results show that even though the models trained

GAN-Based Differential Private Image Privacy Protection Framework for the Internet of Multimedia Things

“…More recent image privacy research focuses on the inherent implicit information of the photos. Tonge et al [ 10 ] explore learning models that can automatically classify the private or public parts in an image using deep neural networks. Yu et al [ 11 ] create a new tool called “iPrivacy” which uses a deep learning algorithm to detect privacy-sensitive objects.…”

“…To achieve this, privacy protection methods need to detect, and then cover/remove/replace sensitive content in images. Several recent studies have explored this direction [ 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 ]. For example, Viola et al [ 8 ] used a sliding window detector to identify and blur the license plates in Google Street View images.…”

Knowledge Distillation with BERT for Image Tag-Based Privacy Prediction

“…With the rapid growth of the number of users on online social networking sites, image privacy has become a major concern (Ahern et al, 2007;Squicciarini et al, 2017). Users may accidentally disclose their sensitive information such as locations, habits or personal relationships from images that they post to their social networking sites (Squicciarini et al, 2017), which could be used in the detriment of the users (Tonge and Caragea, 2020). Zerr et al (2012a) defines private images as ones that belong to the private sphere (e.g., portraits, family, home) or capture sensitive contents that can not be shared with everyone on the Internet.…”

FaceIDP: Face-IDentification Privacy under theCompressed Sensing Framework

“…could combine a user's multiple facial images to form the 3D feature point cloud and breaches the FaceID system to identify the user of interest. This is especially true in the era of Artificial Intelligence (AI): through training large number of facial images of the users, face feature vectors could be learned accurately; then the face identification of the user is carried out through deep learning, leading to privacy leakage from mining information of the publicly shared facial images [4], [5], [6], [7]. Thus the FaceID systems face the real risk of being breached and FaceID is forbidden in many cities, such as San Francisco and Boston, in USA.…”