Intel FakeCatcher Can Detect Deepfake Videos With 96% Accuracy

 


Deepfake technology allows an individual's face to be pasted onto another individual's face in a video. This can lead to defamation if the technology is used for the wrong purpose. To detect this forgery is also not easy because Deepfake technology is maturing. Therefore Intel launched FakeCatcher software which they say has 96% accuracy in detecting fakes in videos.



What makes FakeCatcher interesting is that this artificial intelligence detects blood flow to distinguish genuine videos from fake videos. On a real human face, changes in the skin can be detected as blood flows. This does not happen with faces patched using technology such as DeepFake.



FakeCatcher was developed by Ilke Demi from Intel Labs and Umur Ciftci from the State University of New York at Binghamton.

Previous Post Next Post