Meta’s* new neural network, Movie Gen, enables the creation of realistic videos from text prompts
Users can create videos from scratch or edit existing ones by changing the background, color correction, or adding effects. You can also create a video featuring a specific person by uploading their photo, add any sound effect to the video, or replace an object (for example, turn a Chinese lantern into a soap bubble). One of the key features is the ability to set the aspect ratio.
Movie Gen outputs 10-second videos (up to a maximum of 16 seconds) in Full HD at 24 fps, but the resolution is upscaled from 768×768.
Unfortunately, the Movie Gen neural network is currently only for demonstration purposes and is not available to users.
*Meta is recognized as an extremist organization in Russia.
