Intel’s FakeCatcher deepfake detector uses blood flow to spot fakes

Deepfake - Magnifying Glass

The system spots blood flow in the pixels of a video or image to determine whether or not it is real.

Intel’s FakeCatcher deepfake detector has achieved a 96 percent accuracy rate in real time by using blood flow in the pixels of an image or video to determine if it is real.

Skillfully and believably altered video and images are now all over the internet.

Deepfake videos and images are synthetic media that alters a person’s face, voice, or both in order to generate an entirely new falsified image or video of people or circumstances. Most people have already seen them, and the odds are that they had no idea that what they were seeing wasn’t real.

Deepfake - Fake - Videos

The characteristics of these pieces are very real looking and have contributed to the spread of hoaxes, misinformation, and fraud, disseminating online. Intel has responded through the development of new FakeCatcher technology, which boasts a 96 percent accuracy in detecting fakes over real images and video.

Deepfake videos and images are made through the use of impressive machine learning and AI tech.

The outcomes can be exceptionally accurate impressions of politicians, celebrities and other public figures saying or doing things that simply did not occur in reality. Existing detection tech can require hours to spot the fakes, since they investigate digital manipulation indicators through the use of deep learning.

FakeCatcher is able to detect fake images and videos in real time by “assessing what makes us human – ‘blood flow’ in the pixels of a video,” said a press release issued by Intel about the technology. The company described the tech as able to spot changes in the color of a person’s veins as a result of the blood circulating through the body. Indications of blood flow are collected from the images of a person’s face and the FakeCatcher algorithms then determine whether what is being seen is fake or real.

This type of deepfake detecting technology is becoming increasingly important as it can help individuals and companies to be able to spot false images and videos and to be able to avoid potentially harmful outcomes.

Leave a Comment


This site uses Akismet to reduce spam. Learn how your comment data is processed.