AAAI Publications, Thirtieth AAAI Conference on Artificial Intelligence

Font Size: 
Image Privacy Prediction Using Deep Features
Ashwini Kishore Tonge, Cornelia Caragea

Last modified: 2016-03-05

Abstract


Online image sharing in social media sites such as Facebook, Flickr, and Instagram can lead to unwanted disclosure and privacy violations, when privacy settings are used inappropriately. With the exponential increase in the number of images that are shared online, the development of effective and efficient prediction methods for image privacy settings are highly needed. In this study, we explore deep visual features and deep image tags for image privacy prediction. The results of our experiments show that models trained on deep visual features outperform those trained on SIFT and GIST. The results also show that deep image tags combined with user tags perform best among all tested features.

Keywords


Deep visual feature, Deep tag, User tag, Deep neural network, Deep image tag, Social networking site, Image privacy classification, Neural network, Privacy setting, Deep feature

Full Text: PDF