Image Provenance Analysis for Disinformation Detection
Session Recording and Materials
Join Red Hat Research for the next Research Days event, “Image Provenance Analysis for Disinformation Detection” on July 21, 2022 from 11AM to 12:30PM EDT (5PM CEST, 6PM IDT).
Given the pervasiveness of image and video data in society, being able to identify when images have been altered is very important. This discussion will provide insights into scalable methods for analyzing media and detecting altered images, and explore some potential larger societal impacts of this work.
Walter Scheirer, Dennis O. Doughty Collegiate Associate Professor, University of Notre Dame and Daniel Moreira, Incoming Assistant Professor, Loyola University will discuss their work in provenance analysis and their efforts to extend it to unveiling disinformation campaigns. Jason Schlessman, Principal Software Engineer, Red Hat, will lead the conversation, which attendees are welcome to join live.
Composite images are the outcome of combining pieces extracted from two or more other images, sometimes with the intent to deceive the observer and convey false narratives. Consider an image suspected of being a composite, and a large corpus of images that might have donated pieces to the composite (such as photos from social media). Inside media forensics, provenance analysis is the problem of (1) finding, within the available corpus, the images that either directly or transitively share content with the composite (namely, the task of provenance retrieval), as well as of (2) establishing the directed acyclic graph whose nodes individually represent the composite and related images, and whose edges express the derivation and content-donation story (e.g., cropping, blurring, splicing) between pairs of images, linking seminal to derived elements (namely, the task of provenance graph construction). In this conversation, we will discuss our most recent advances in provenance analysis, concluding with our latest endeavors towards extending it to unveil disinformation campaigns.
Jason Schlessman, Principal Software Engineer, Red Hat
Affiliated Graduate Student
Bill Theisen, PhD student in the Computer Vision Research Lab at the University of Notre Dame