Runway AI introduces new image generation model Frames
Runway AI, a New York-based AI research and technology firm, has introduced a new image generation model called Frames, which marks "a big step forward in stylistic control and visual fidelity."
The model appears to be specialized for stylistic consistency, i.e., generating the same type of image with some new details added or removed, something other image generation models currently struggle with.
Runway says that Frames allows you to pick a stylistic choice for your project "and reliably generate variations that stay true to your aesthetic."
Runway showed off some examples of stylistic consistency with Frames in an X thread. The stylistic choices referred to as "Worlds" range from 1980's SFX, 1970s album art, Japanese Zine, realistic digital pictures, landscapes, and much more.
The images generated look impressive at first glance and maintain consistency in theme and vision. Runway says it will gradually roll out access to Frames inside the Gen-3 Alpha of its video generation suite and Runway API.
Runway's video generation tools have been utilized in films like Everything Everywhere All At Once and in music videos for artists like Kanye West. They have even been used in television shows like Top Gear.
