AI Startup Runway claims that his latest AI video model can generate coherent scenes and people in many shots, As announced. Movies generated by AI may have a problem maintaining a coherent story, but the catwalk Claims regarding x That the novel model, Gen-4, should enable users to be more “continuity and control” by telling stories.
The novel Gen-4 video synthesis model, which would currently implement paid and corporate users, enables users to generate characters and objects in shots using one reference image. Users must then describe the desired composition, and then the model will generate consistent results from many sides.
As an example, the startup has released a video from a woman keeping her appearance in various shots and contexts in various lighting conditions.
The edition appears less than a year after the announcement of the catwalk Gen-3 alpha video generator. This model has expanded the length of films that can be produced by users, but also caused controversy, because apparently it was trained for thousands of scraped videos from YouTube and pirate films.