The name of the project is “learning from videos”, and aims to improve learning systems to give life to new applications and create more precise viewing recommendations for subscribers
Use artificial intelligence to better monitor and catalog video content on Facebook, to improve learning systems, create new applications, monitor content more carefully and offer more timely viewing recommendations for users. This is the goal of “Learning from video”, the new Facebook project aimed at using AI to understand video content.
“By learning from global streams of publicly available video, for hundreds of languages, our AI systems will not only improve accuracy, but also adapt to a rapidly changing world, to recognize the nuances and visual cues of different cultures and regions – reads a post on the Facebook corporate blog – By helping AI researchers to free themselves from addiction to labeled data, we can create completely new experiences”.
Among the first services that will be able to make use of this project, the company has selected Reels, the social network created by Menlo Park following the example of TikTok: thanks to Learning from video the app will improve the advice for users, with the aim of involving them. More in the activities that will be more interesting depending on the profiles of the members.
The artificial intelligence system pioneered by Facebook will be able to analyze not only images, but also audio and textual inputs. What the social media founded by Mark Zuckerberg has not yet specified is whether this system will sooner or later also be used for advertising targeting.
