© QubixStudio/Shutterstock.com
The Meta group, of which Facebook is a member, continues to show its progress in the field of artificial intelligence. As a reminder, it already offers an equivalent of ChatGPT that is based on its Llama AI models. And like OpenAI, Meta is also interested in other formats, including video. As the group recalls in a blog post, it has already presented artificial intelligences capable of generating images, videos, 3D animations, sound, or even editing images. And a few days ago, it unveiled Movie Gen, a tool that combines all of these features, while allowing for more control over the generated video.
“Our latest research shows how you can use simple text input to produce custom videos and sounds, edit existing videos, and transform your personal image into a unique video,” Meta explains. The group’s goal is to make video creation accessible to more people, while specifying that Movie Gen is not intended to replace the work of artists and animators. “We are sharing this research because we believe in the power of this technology to help people express themselves in new ways and to provide opportunities to people who otherwise would not have them,”, the group says.
🎥 Today we’re premiering Meta Movie Gen: the most advanced media foundation models to-date.
Developed by AI research teams at Meta, Movie Gen delivers state-of-the-art results across a range of capabilities. We’re excited for the potential of this line of research to usher in… pic.twitter.com/NDOnyKOOyq
— AI at Meta (@AIatMeta) October 4, 2024
A complete tool
First of all, from a description provided by the user, Movie Gen is able to generate videos of 16 seconds, maximum, with a refresh rate of 16 frames per second.
It is also possible to send the image of a person to this new AI and ask it to generate a video from this image.
In addition, Movie Gen allows also, using prompts, to edit an existing video. Meta specifies that the AI only modifies the pixels concerned by the edition.
200% Deposit Bonus up to €3,000 180% First Deposit Bonus up to $20,000Furthermore, if a video has no sound, Movie Gen can, for example, create background noise. And the user can give instructions by including text in their prompt. Meta indicates that the generated audio files can last up to 45 seconds and are of “high quality” and “high fidelity”.
More examples of what Meta Movie Gen can do across video generation, precise video editing, personalized video generation and audio generation. pic.twitter.com/Dz65QGNgaJ
— AI at Meta (@AIatMeta) October 4, 2024
Meta challenges the creator of ChatGPT
In any case, Meta believes that Movie Gen is currently the best AI of its kind. This claim is based on human evaluations. These evaluations reportedly revealed that Movie Gen is more efficient than Sora, OpenAI's video generation artificial intelligence.
But Meta isn't offering the AI to the general public right now. It says it's continuing to improve Movie Gen by working with filmmakers and creators. “By taking a collaborative approach, we want to make sure we’re creating tools that help people enhance their inherent creativity in new ways they may never have dreamed of,”, Meta says.
In any case, he’s already imagining the potential uses of Movie Gen on his products, such as Facebook and Instagram’s Réels videos, or WhatsApp.
- In addition to having developed a competitor to ChatGPT, based on his Llama models, Meta is also interested in generating images, videos and sounds
- This one has just lifted the veil on Movie Gen, a new AI that can generate videos, create videos from a person's image, edit existing videos, or generate sound for a video that doesn't have any
- Meta's goal is to make video creation more accessible and it is already imagining how it could integrate Movie Gen into Facebook, Instagram or WhatsApp
- But for the moment, Movie Gen is not yet offered to the general public
📍 To not miss any Presse-citron news, follow us on Google News and WhatsApp.
155.0 M reviews
[ ]