Facebook has announced a new AI project that can learn from the visual, auditory and textual content of videos. This project, called Learning from Videos, is an important step forward for the machines of the future that can ‘learn like humans’.
Artificial intelligence algorithms are now with us on the platforms we visit every day, on the phones we receive, in short, at every moment of our lives. This means that every major innovation depends mainly on advances in artificial intelligence algorithms.
That’s why all the tech giants are working for their more ‘intelligent’ AI, which can learn better every time. However, at this point, the way an AI is learning is quite different from the way a person learns. But in the future, it is thought that artificial intelligence will become more and more human-like. A new project announced by Facebook is an important step for machines that can learn like humans in the future.
An artificial intelligence that can learn every aspect of video:
Underneath the basic principle of operation of all the platforms that Facebook incorporates are artificial intelligence algorithms. All the ads and suggestions we come across are powered by artificial intelligence. Facebook’s new project, Learning from Videos, will also provide its users with more consistent and relevant recommendations, while providing a kind of introduction to artificial intelligence to be used in human-like learning machines of the future.
Learning from Videos algorithms will be able to learn audio, visual elements of all videos uploaded to Facebook platforms. The first application of the new project started to be used on Instagram Reels videos. This will bring users more consistent videos based on their interests and favorite content.
Learning from videos is a challenging task for AI because certain situations, such as noise pollution in a video, can make learning difficult. But there are various studies in this field, and Facebook has been using similar algorithms for some time. For example, Facebook has achieved successful results in detecting hate speech in content through sound.
The Learning from Videos algorithm will be available in all countries for all languages. It will also be an important starting point for AI machines with further learning capabilities that will come up in the future.