Presentation
Comparing Human to Analytic Performance on Detecting and Characterizing Manipulated Media
DescriptionEmerging digital technologies and the ubiquity of social media platforms have accelerated the spread of manipulated media, news, and technical information in both size and scale. As media manipulations increase in technical sophistication and semantic complexity, efforts to improve analytic performance are increasingly challenging. The Semantic Forensics (SemaFor) program—led by the Defense Advanced Research Projects Agency (DARPA)—aims to develop foundational technology to refine and extend methods for detecting, attributing, and characterizing semantic inconsistencies in falsified media at scale through a series of evolving tasks focusing on areas of interest to the community. The goal of the present research is to provide a baseline of human performance for comparison with analytics developed to address detection and characterization demands across diverse manipulated media tasks.
Contributors
Principal Research Engineer and Deputy Division Director
Event Type
Lecture
TimeWednesday, September 11th1:50pm - 2:10pm MST
LocationFlagstaff
Human AI Robot Teaming (AI)