To reduce fake photos and videos professionally

Facebook uses artificial intelligence to protect its customers from "fakes"

Counterfeits are a problem for websites when they are used to manipulate Internet dealers.

archival

Two Facebook researchers have come up with an AI-based method that detects "deepfakes" — images and videos that are highly crafted to help identify their provenance.

Such forgeries pose a problem on websites, as they may be used to manipulate Internet users or discredit people by installing clips that show them saying or doing things they did not really say or do.

These synthesis processes are based on artificial intelligence techniques.

"Our system will make it easier to detect deepfakes and track information associated with them," said Facebook researchers Tal Hasner and Shi Yin, who worked on the topic with Michigan State University.

The method reported by the researchers provides "tools to facilitate the investigation of events associated with coordinated disinformation operations using deepfakes."

To develop this system, the researchers used a technique called "reverse engineering", which is based on deconstructing the way a product is made, in this case a video or image.

The software used by the system detects any gaps in the synthesis process (montage), affecting the digital footprint of the images.

In the field of photography, this fingerprint allows identification of the camera used.

In informatics, this technology can "recognize the system used to make the counterfeit operations", according to the researchers.

Follow our latest local and sports news and the latest political and economic developments via Google news