Author:
Moreira Daniel,Theisen William,Scheirer Walter,Bharati Aparna,Brogan Joel,Rocha Anderson
Abstract
AbstractThe literature of multimedia forensics is mainly dedicated to the analysis of single assets (such as sole image or video files), aiming at individually assessing their authenticity. Different from this,image provenance analysisis devoted to the joint examination of multiple assets, intending to ascertain their history of edits, by evaluating pairwise relationships. Each relationship, thus, expresses the probability of one asset giving rise to the other, through either global or local operations, such as data compression, resizing, color-space modifications, content blurring, and content splicing. The principled combination of these relationships unveils the provenance of the assets, also constituting an important forensic tool for authenticity verification. This chapter introduces the problem of provenance analysis, discussing its importance and delving into the state-of-the-art techniques to solve it.
Reference64 articles.
1. Advance Publications, Inc (2012) Reddit PhotoshopBattles. https://bit.ly/3ty7pJr. Accessed on 19 Apr 2021
2. Amazon Mechanical Turk, Inc (2021) Amazon Mechanical Turk. https://www.mturk.com/. Accessed 11 Apr 2021
3. Baeza-Yates R, Ribeiro-Neto B (1999) Modern information retrieval, vol 463, Chap 8. ACM Press, New York, pp 191–227
4. Ballard D (1981) Generalizing the Hough transform to detect arbitrary shapes. Elsevier Pattern Recognit 13(2):111–122
5. Bay H, Ess A, Tuytelaars T, Van Gool L (2008) Speeded-up robust features (SURF). Elsevier Comput Vis Image Understand 110(3):346–359
Cited by
1 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献
1. SynthProv: Interpretable Framework for Profiling Identity Leakage;2024 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV);2024-01-03