Meta’s Film Gen AI Mannequin Can Generate Video With Sound or Music

Meta’s Film Gen AI Mannequin Can Generate Video With Sound or Music

Fb proprietor Meta introduced on Friday it had constructed a brand new AI mannequin referred to as Film Gen that may create realistic-seeming video and audio clips in response to consumer prompts, claiming it will possibly rival instruments from main media era startups like OpenAI and ElevenLabs.

Samples of Film Gen’s creations offered by Meta confirmed movies of animals swimming and browsing, in addition to movies utilizing individuals’s actual photographs to depict them performing actions like portray on a canvas.

Film Gen can also generate background music and sound results synced to the content material of the movies, Meta mentioned in a weblog submit, and use the software to edit present movies.

In a single such video, Meta had the software insert pom-poms into the palms of a person working by himself within the desert, whereas in one other it modified a parking zone the place a person was skateboarding from dry floor into one coated by a splashing puddle.

Meta’s Film Gen AI mannequin lets customers edit movies utilizing textual content prompts
Photograph Credit score: Meta

 

Movies created by Film Gen will be as much as 16 seconds lengthy, whereas audio will be as much as 45 seconds lengthy, Meta mentioned. It shared information displaying blind assessments indicating that the mannequin performs favorably in contrast with choices from startups together with Runway, OpenAI, ElevenLabs and Kling.

The announcement comes as Hollywood has been wrestling with the best way to harness generative AI video expertise this yr, after Microsoft-backed OpenAI in February first confirmed off how its product Sora may create characteristic film-like movies in response to textual content prompts.

Technologists within the leisure business are keen to make use of such instruments to reinforce and expedite filmmaking, whereas others fear about embracing methods that seem to have been educated on copyright works with out permission.

Lawmakers even have highlighted issues about how AI-generated fakes, or deepfakes, are being utilized in elections world wide, together with within the U.S., Pakistan, India and Indonesia.

Meta spokespeople mentioned the corporate was unlikely to launch Film Gen for open use by builders, because it has with its Llama sequence of large-language fashions, saying it considers the dangers individually for every mannequin. They declined to touch upon Meta’s evaluation for Film Gen particularly.

As an alternative, they mentioned, Meta was working straight with the leisure group and different content material creators on makes use of of Film Gen and would incorporate it into Meta’s personal merchandise someday subsequent yr.

In accordance with the weblog submit and a analysis paper in regards to the software launched by Meta, the corporate used a mixture of licensed and publicly accessible datasets to construct Film Gen.

OpenAI has been assembly with Hollywood executives and brokers this yr to debate potential partnerships involving Sora, though no offers have been reported to have come out of these talks but. Anxieties over the corporate’s strategy elevated in Could when actor Scarlett Johansson accused the ChatGPT maker of imitating her voice with out permission for its chatbot.

Lions Gate Leisure, the corporate behind “The Starvation Video games” and “Twilight,” introduced in September that it was giving AI startup Runway entry to its movie and tv library to coach an AI mannequin. In return, it mentioned, the studio and its filmmakers can use the mannequin to reinforce their work.

© Thomson Reuters 2024

(This story has not been edited by NDTV employees and is auto-generated from a syndicated feed.)