Intel has revealed its latest project that can detect Deepfakes with 96% accuracy. Called FakeCatcher, Intel's new detection technology can flag Deepfakes within milliseconds.
Deepfakes have gained notoriety as being part of a new synthetic medium that's potential uses can possibly hurt a person's public image. They have unfortunately been commonly associated with various pornographic material, fake news and even financial fraud.
The major problem is how easily available the tools for creating Deepfakes are, and they have slowly spread to infiltrate all mediums of entertainment.
Deepfakes use machine learning algorithms and AI to convincingly emulate a person's likeness in order to deceive. The potential uses for this technology are varied and dangerous.
A recent study conducted by PNAS stated that the technology can emulate faces so convincingly that they are almost indistinguishable for the real ones.
Intel's FakeCatcher attempts to solve the problem of recognizing faked media by using new real-time detection techniques that use an interesting method to identify fakes.
“When our hearts pump blood, our veins change color. These blood flow signals are collected from all over the face and algorithms translate these signals into spatiotemporal maps. Then, using deep learning, we can instantly detect whether a video is real or fake,” said Intel in a press release.
The technology runs on servers powered by Intel hardware and software and the company says that the platform, "can run up to 72 different detection streams simultaneously on 3rd Gen Intel Xeon Scalable processors".
Discover the latest Business News, Sensex, and Nifty updates. Obtain Personal Finance insights, tax queries, and expert opinions on Moneycontrol or download the Moneycontrol App to stay updated!
