Most of the time, offensive images, videos, and text is usually caught after it has been posted, and sometimes it does require intervention on our part as sometimes users will have to report the offending media. However Facebook is hoping to change this with the use of AI in which it will hopefully be able to identify and flag offensive live videos.
Facebook Live is one of the company’s latest features that is gaining in popularity as all kinds of people are taking advantage of it, whether it be by celebrities to chat with their fans, or news outlets to share live reportings, or just for fun by regular users to show their friends where they are in real-time. However we expect that it is possible that the tool could be abused in all kinds of ways, so for Facebook to police it manually would be a monumental task.
According to Joaquin Candela, the company’s director of applied machine learning, this is where AI comes in. Candela says that this AI is “an algorithm that detects nudity, violence, or any of the things that are not according to our policies,” and will then flag videos that it detects might contain those things.
The AI has already been in testing and Facebook has apparently been using it to flag extremist video content, and now it is being applied to live videos as well. However Candela explains that there are challenges with this application. “Your computer vision algorithm has to be fast, and I think we can push there. The other one is you need to prioritize things in the right way so that a human looks at it, an expert who understands our policies, and takes it down.” We suppose it’s still a good start.