As it is quite evident that Deepfake AI is all over the news (for the wrong reasons), we know that the implications are serious. Hence, it’s time we got ways to fight it, and that’s what Facebook and many more companies are trying to do.
Facebook’s Deepfake Detection Challenge
Facebook has announced its Deepfake Detection Challenge to come up with ways to identify and fight the Deepfake technology used to manipulate videos and photos.
The initiative will be in collaboration with Microsoft and universities such as Cornell Tech, MIT, University of Oxford, UC Berkeley, University of Maryland, College Park, and University at Albany-SUNY.
How will the ‘Facebook Deepfake Challenge’ work?
Facebook’s challenge will involve new data sets (new Deepfake videos featuring paid actors made explicitly for this purpose), grants, and prizes.
Yes, you hear it right! Facebook has allocated $10 million to the challenge, and those who can identify the Deepfake videos will be awarded the prize money.
Another thing worth noting is that Facebook won’t use its users’ data as data sets for the challenge, which comes as a sigh of relief.
Find your dream job
The data sets will be tested via a targeted technical working session at the International Conference on Computer Vision (ICCV) in October 2019.
Eventually, the data sets will be released at the Conference on Neural Information Processing Systems (NeurIPS) this December.
Facebook’s efforts to fight fake news
While the initiative is not built with a lot of precision, the fact that Facebook is making some (to clear its Karma, of course!) can’t be dismissed.
Why do we need companies to do something?
The idea of the Deepfake Detection Challenge is to easily detect Deepfake content and stop the manipulation of media, which will eventually curb the spread of fake news and public shaming of people.
The spread of fake news has been a non-stop process, and deepfake adds to its strength. We need concrete methods to fight it, and tech companies’ efforts can help us big time.
It’s time for us to realize that Deepfakes are not really a fun element (I know we like those Christopher Nolan fake videos) but also a source of malicious activities, intrusion of privacy, and a lot more.
As it’s the initial stage of companies trying to come together to find a solution, we have to wait until something concrete and worthy is made. So keep visiting us for all the tech news you need!