AI startup Runway says its newest AI video mannequin can generate constant scenes and folks throughout a number of photographs, according to an announcement. AI-generated movies can battle with sustaining constant storytelling, however Runway claims on X that the brand new mannequin, Gen-4, ought to permit customers extra “continuity and management” whereas telling tales.
At the moment rolling out to paid and enterprise customers, the brand new Gen-4 video synthesis mannequin permits customers to generate characters and objects throughout photographs utilizing a single reference picture. Customers should then describe the composition they need, and the mannequin will then generate constant outputs from a number of angles.
For instance, the startup launched a video of a lady sustaining her look in numerous photographs and contexts throughout a wide range of lighting circumstances.
The discharge comes lower than a yr after Runway introduced its Gen-3 Alpha video generator. That mannequin prolonged the size of movies customers may produce, however also sparked controversy as reportedly had been educated on hundreds of scraped YouTube movies and pirated movies.