Image origin identification for online social networks (OSNs)
Sun W.; Zhou J.
Source PublicationProceedings - 9th Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2017
AbstractThe rapid development of various online social networks (OSNs) makes it unprecedentedly popular to share photos online. This leads OSNs one of the main provenances of online images. However, the illegal activities on these online contents such as misusing and dissemination widely exist. Under this circumstance, the identification of the origin and the propagation path of an online image is crucial for many forensic applications. In this work, we propose a simple yet effective method to determine the image origin by exploiting the unique traces left by the operations of different OSNs. To this end, we first conduct a comprehensive study on the manipulations that various OSNs perform on uploaded images. Based on the knowledge of these operations, we design a feature vector and eventually train a SVM classifier for identifying where these online images come from. Extensive experimental results are provided, showing that the proposed method achieves very high accuracy of image origin identification, and outperforms the state-of-the-art work.
URLView the original
Fulltext Access
Citation statistics
Document TypeConference paper
CollectionUniversity of Macau
AffiliationUniversidade de Macau
Recommended Citation
GB/T 7714
Sun W.,Zhou J.. Image origin identification for online social networks (OSNs)[C],2018:1512-1515.
Files in This Item:
There are no files associated with this item.
Related Services
Recommend this item
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[Sun W.]'s Articles
[Zhou J.]'s Articles
Baidu academic
Similar articles in Baidu academic
[Sun W.]'s Articles
[Zhou J.]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[Sun W.]'s Articles
[Zhou J.]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.