Adobe has introduced the Firefly Video Model, the most recent addition to their Firefly family of generative AI services. This new application allows users to create videos using basic text prompts and pictures. While the Firefly Video Model is planned to be available in beta by the end of the year, Adobe is already updating their popular video-editing program, Premiere Pro, with Generative Extend, a feature that expands the canvas of video frames by producing new backgrounds.
The Firefly Video Model expands Adobe's generative AI capabilities by putting text-to-video and image-to-video functionalities in the hands of video editors and producers. Using Adobe's proprietary AI technology, video editors may easily alter video features like camera angles, motion and zoom based on text prompts, as well as generate B-roll material from reference photos.
Along with creating video from text, the model's image-to-video capability can turn still photographs and illustrations into live-action videos. This will allow creators to experiment with new video production techniques, streamline workflows and increase creative ideas.
Ashley Still, Senior Vice President of Adobe's Creative Product Group, stated, "Building upon our foundational Firefly models for imaging, design and vector creation, our Firefly foundation video model is designed to help the professional video community unlock new possibilities, streamline workflows and support their creative ideation."
As part of Adobe's commitment to responsible AI, the company stated that the Firefly Video Model is only trained on content that it has permission to use, ensuring that it is financially safe and respects intellectual property. "The Firefly Video Model is designed to be commercially safe and is only trained on content we have permission to use, never on Adobe customer content," the company said.
Adobe's Firefly Video Model is not restricted to simple video creation. It can also be used for more complicated jobs like producing animations with fine control over camera angles and motion, as well as inserts with motion to improve important video shots. Users can even combine AI-generated video content with real-world footage.