The Cyber Security Review | Tuesday, November 29, 2022
Intel's FakeCatcher detects a deepfake in real time with a 96 per cent accuracy rate.
FREMONT, CA: Deepfakes, or synthetic media that alter an image or video by using another person's voice or face to create a new, false representation of a person or event, have proliferated online in recent years. Deepfakes are extremely convincing features that have made it possible for many internet hoaxes, frauds, and misinformation to spread. In response, Intel (NASDAQ: INTC) unveiled FakeCatcher, a new tool with a 96 per cent accuracy rate for detecting deepfake media.
Stay ahead of the industry with exclusive feature stories on the top companies, expert insights and the latest news delivered straight to your inbox. Subscribe today.
Deepfakes are frighteningly accurate impersonations of politicians and celebrities doing and saying things they haven't using remarkable technology generated from machine learning and artificial intelligence. Existing solutions utilise deep learning to look for indicators of digital tampering, which can take hours to uncover in deepfakes. According to a news release, Intel's FakeCatcher can spot a deepfake in real time by evaluating what makes human blood flow' in the pixels of a video.
According to Intel, its technology can detect alterations in the colour of veins caused by blood flow. Afterwards, algorithms translate blood flow signals from the face to determine whether a video is authentic or a deepfake.
To prevent negative effects, it is becoming more and more crucial to have tools that can assist in recognising deepfakes. Deepfake films and images can range from being graphic to spreading misinformation about the media.
Scammers have already employed deepfakes to pretend to be job seekers to access private company data. They've also been employed to make insulting statements by posing as well-known political personalities.
Most people mindlessly scroll through their Twitter feed and don't take the time to determine whether a video is real or fake, even though some movements and mannerisms in deepfakes give away their deceptive nature.
And by the time a deepfake receives millions of shares, it is already too late since, according to the Pew Research Center, 63 per cent of US respondents say that a changed video has confused them about current events.
More in News