In recent years, more and more deepfakes abound on the internet. These are digital files – photos or videos – doctored. Their authors take an image or video and use another person’s face or voice to create a new false image of people or events. It is therefore very likely that you have already seen one without knowing that it was a deepfake.
The very real appearance of deepfakes has allowed many cases of misinformation and hoaxes to spread online. In response, Intel announced a new technology called “FakeCatcher” to detect deepfake media with an accuracy rate of 96%.
Deepfakes use impressive technology derived from machine learning and artificial intelligence to create chillingly accurate images and videos of celebrities and politicians doing and saying things they didn’t say.
Existing technologies can take hours to dispel internet users’ trust in a deepfake
Existing technologies can take hours to dispel internet users’ faith in a deepfake, as they use deep learning to study signs of digital manipulation.
According to a press release from the company, FakeCatcher can detect a deepfake in real time by “assessing what makes us human – the ‘blood flow’ in the pixels of a video”.
Intel says its technology can identify the color changes in our veins as blood flows through the body. Blood flow signals are then collected from the face and translated by algorithms to discern whether a video is real or a deepfake.
Most people don’t take the time to check if a video is real or fake.
It is increasingly important to have software to help us identify deepfakes.
In the past, scammers have used deepfakes to impersonate job seekers to gain access to sensitive company information. They have also been used to impersonate prominent political figures in order to make inflammatory remarks.
Especially since most people mindlessly scroll through their Twitter feed and don’t take the time to check if a video is real or fake. And when a deepfake gets millions of shares, it’s way too late.
Source: ZDNet.com