Since March Facebook had been researching and testing with 27
fact-checking partners in 17 countries around the world how to fact check false photos and videos. They are regularly onboarding new partners.
They have determined that misinformation in photos and videos usually falls into three categories: (1) Manipulated or Fabricated, (2) Out of Context, and (3) Text or Audio Claim. These are the kinds of false photos and videos regularly seen on Facebook. They hope to reduce the frequency of such imagery by expanding photo and video fact-checking.
They have built a machine learning model that uses various engagement signals, including feedback from people on Facebook, to identify potentially false content. Then they send those photos and videos to fact-checkers for their review. Fact-checkers can also surface content on their own.
“Many of our third-party fact-checking partners have expertise evaluating photos and videos and are trained in visual verification techniques, such as reverse image searching and analyzing image metadata, like when and where the photo or video was taken. Fact-checkers are able to assess the truth or falsity of a photo or video by combining these skills with other journalistic practices, like using research from experts, academics or government agencies,” according to Facebook.
For more about the process check out this
link. For more information about how they understand the text in images and videos with machine learning check out this
explanation.