Goodbye to DeepFake: Intel develops AI to detect counterfeiting in networks

Intel has introduced a new technology based on deepfake detection algorithms .


Intel is speeding up on issues parallel to processors and graphics cards, and is aiming for fast information management by implementing artificial intelligence for rapid data recognition.

Now, and as part of its social responsibility work, the semiconductor brand presents FakeCatcher, a technology capable of detecting 96% deepfake videos .

In a statement , Intel points out that it is important to give the world a deep fake detection tool: “Deepfake videos are everywhere now. You’ve probably already seen them; videos of celebrities doing or saying things they never actually did,” says Ilke Demir, a senior research scientist on the Intel Labs staff.

Through face detection algorithms optimized by multi-threaded software libraries applied to real-time video and image processing tools, Intel was able to analyze “blood streams,” elements that make us human against “computing eyes.” .

The function of FakeCatcher, one of the first of its kind, is to detect these subtleties in human faces, and treat them on spatiotemporal maps to detect, through deep learning, if a video is fake or not.

DeepFakes: A growing threat in the world

The Intel report includes figures from Gartner that warn of a total spending of 188 billion dollars on cybersecurity solutions that prevent the malicious entry of deepfake tools into digital environments. In simple terms, the detection process is currently expensive and time consuming.

According to the company, the inclusion of FakeCatcher would help social media platforms in blocking potentially harmful videos, just as global news agencies would be able to detect fake videos and thereby prevent the publication of unverified news.