As part of its Responsible AI Initiative, semiconductor chip manufacturer Intel has introduced a new technology dubbed FakeCatcher that can detect deepfakes in real-time. According to the company's representatives, FakeCatcher can do that with a 96% accuracy rate.
Deepfakes are fake images or videos of real people made using AI technology. They are becoming easier to create and more widely used for spreading disinformation, so it is essential to have a tool that can detect and fight them. The newly developed technology uses Intel's OpenVINO (an open-source deep learning toolkit) “to run AI models for face and landmark detection algorithms.”
FakeCatcher analyzes subtle facial blood flows in video pixels to determine if a video subject is real or fake. The new deepfake detector collects blood flow signals from all over the face and translates them into spatiotemporal maps. It then uses deep learning to determine whether a video is real or fake.
However, there's no information about the timing or a probability of a public release just yet.