New Algorithm Developed to Protect Against Deepfakes

As deepfake technology continues to advance, it is becoming much more difficult to verify the authenticity of information published online. These technologies make it easy to manipulate images and videos using neural networks capable of simulating speech, movements, and the appearance of people.

In response to this challenge, a group of researchers from Boston University has developed an algorithm designed to protect images from being used as the basis for deepfakes. The algorithm applies an invisible filter to images or videos. If someone tries to use a neural network to modify these files, the result will either remain unchanged or become completely distorted. The filter alters the pixels in such a way that the videos and images become unusable for creating deepfakes.

The project developers have made the source code for this technology available on GitHub and have also published a video demonstrating how it works.

Source

• Our other channels
• Our friends and partners