Adobe said on Wednesday that its AI video-generation model will launch towards the end of the year. Just as it did with images, Adobe aims to offer a tool for generating videos based solely on licensed content.
According to the company, Firefly's AI model for videos is designed to create cinematic videos that professional editors and creators can use as inspiration or to fill gaps in a project and add new elements to existing video.
The tool will be available initially on the Adobe Firefly website and will feature text-to-video and image-to-video options. According to Adobe, a wide range of controls such as changing the camera angle or movement will be offered.
The company is also developing a functionality called “Generative Extend” that will be integrated directly into Premier Pro.
Adobe's chief technology officer for digital media, Ely Greenfield, said Adobe's approach to the video model remains the same as the image generation model, where copyright and content creators will be respected.
“What sets the Adobe Firefly video model introduced today apart from other AI-powered video generation models is that it is designed to be commercially safe and trained only on content we have permission to use – and never on Adobe customer content,” said Greenfield.
edition
t7 live
