DeepFake-o-meter: exposing DeepFakes

DeepFakes refer to AI-synthesized fake media. By creating illusions of an individual's presence and activities that did not occur in reality, DeepFakes can cause real harms when they are weaponized. DeepFake-o-meter is an open-source and user-friendly online platform developed by the UB Media Forensics Lab (UB MDFL) to detect third-party DeepFake algorithms. For users, it provides a convenient service to analyze DeepFake media with multiple state-of-the-art detection algorithms, with secure and private delivery of the analysis result.  For developers of DeepFake detection algorithms, it provides an API architecture to wrap individual algorithms and run on a remote machine. For researchers of digital media forensic, it is an evaluation/benchmarking platform to compare performance of multiple algorithms on the same input.

Director:  Siwei Lyu 

Collaborator:
Dr. Yuezun Li (former student/post-doc), Ocean University of China

or visit the UB Media Forensics Lab.