Intel introduced FakeCatcher, which they say is the first real-time deepfake detector in video. The company claims the technology is 96% accurate and works by analyzing the subtle “blood flow” in video pixels to return results in milliseconds. The technology uses Intel hardware and software, runs on a server, and interacts through a web platform.

Unlike most deep learning-based fake detectors that analyze raw data to pinpoint inauthenticity, FakeCatcher focuses on clues in real videos. It is based on photoplethysmography (PPG), a method of measuring the amount of light that is absorbed or reflected by blood vessels in living tissue. When the heart pumps blood, it enters the veins, which change color.

“You cannot see it with your eyes, but it is computationally visible,” explains Ilke Demir, a senior scientist at Intel Labs who developed FakeCatcher in collaboration with Umur Ciftci of the State University of New York at Binghamton. “PPG signals have been known, but they have not been applied to the deepfake problem before.”

With FakeCatcher, PPG signals are collected from 32 locations on the face, and PPG maps are then created from the temporal and spectral components.

According to a recent research of Eric Horvitz, chief research officer at Microsoft Corporation, the detection of deepfakes is becoming more important with the emergence of threats associated with them. Threats include interactive deepfakes, which create the illusion of talking to a real person, and composite deepfakes, where attackers create many deepfakes to make up a “synthetic story.”

And in 2020, Forrester Research predicted that the costs associated with deepfake fraud will exceed $250 million.

Recently, news about fakes related to famous people has become more frequent. The Wall Street Journal reported on the unauthorized appearance in advertising of fakes of Tom Cruise, Elon Musk and Leonardo DiCaprio, as well as rumor that Bruce Willis gave up the rights to his fake copy, which is not true.