Meta Unveils AI Tool Capable of Generating Audio-Visual Clips
New York, October 4 (Reuters) – Meta, the parent company of Facebook, announced today, Friday, that it has developed a new artificial intelligence model called "MovieGen" capable of generating video clips accompanied by audio to appear realistic based on user requests.
Meta stated that this model can compete with tools from leading media creation startups such as OpenAI and Eleven Labs.
Samples of MovieGen's production, revealed by Meta, included video clips of animals swimming and floating, as well as others using real images of people, showing them engaged in activities like painting on canvas.
In a blog post, Meta said that MovieGen can also generate background music and sound effects synchronized with the content of the video, and the tool can be used to make edits to the clips.
In one of the clips produced by MovieGen, Meta asked the tool to place decorative balls in the hands of a man running alone in the desert. In another clip, it transformed a parking lot where a man was skating on dry ground into one covered by a pool of water.
Meta mentioned that the length of a video clip produced by MovieGen can reach up to 16 seconds, while the audio clip can last up to 45 seconds.
