Mark Zuckerberg, CEO and founder of Meta, revealed today (Thursday) a new system based on artificial intelligence called Make-A-Video. The system turns texts into short and unique video clips.
As mentioned, the new development is based on deep AI capabilities used by image generators, such as Make-A-Scene – a technology developed by Meta’s artificial intelligence teams in Israel, which turns text into images.
To build Make-A-Video, the company’s AI teams worked on another layer of unsupervised learning, which knows how to understand movement in physical space and use it to refine existing image generator technologies.