Meta, the parent company of Facebook, launches 'Make-A-Video' a new system that will allow people to give text information which is then transformed into short, high-quality videos featuring colors, characters and landscapes.
It is based on generative artificial intelligence, one of the emerging applications in the field of AI that applies to the multimedia industry, art and design.
"The videos were generated by an artificial intelligence system created by our Meta team - explains Mark Zuckerberg in a post - You give them a text description and create a video for you. It's a surprising progress, it's much more difficult to generate videos. compared to photos because in addition to correctly generating each pixel, the system must also predict how they will change over time. Make-A-Video solves this problem. We hope to have a beta version that more people can try out soon. "
"Generative AI research is spurring creativity and empowering people to create new content quickly and easily," adds Meta.
Make-A-Video follows the announcement earlier this year of Make-A-Scene, a multimodal generative AI method that gives people greater control over AI-generated content.
"With Make-A-Scene we have demonstrated how people can create photorealistic illustrations and narrative-quality art using words, lines of text and freehand drawing - concludes the parent company of Facebook - We want to ponder how to build new generative AI systems like this one. We are openly sharing research with the community for feedback, we will continue to refine and evolve our approach to this emerging technology. "
