Adobe Premiere Pro Gets Smarter With Firefly: Add Objects, ‘Generative Extend’
Image courtesy of Adobe
After mastering text-to-image, Adobe is pouring its generative prowess into video. The creative tech giant is adding generative AI tools to Premiere Pro, enabling users to perform complex editing tasks with greater ease and precision in hopes of transforming video production workflows.
Powered by Adobe’s Firefly model, these new tools, coming later this year, will allow editors to add or remove objects in real-time, extend footage effortlessly, and maintain transparency with AI-assisted edits.
Object addition is now more intuitive, with editors able to insert or alter objects within their footage using simple text prompts. For example, in a video demo, Adobe showcases how easily realistic diamonds can be added into a scene.
Image courtesy of Adobe
Conversely, the object removal tool allows for the swift deletion of unwanted elements—be it props, crew members, logos, or other distractions—across various frames. In its video, Adobe removes a utility box from a scene, cleaning up the visual space without compromising the flow of the footage.
Image courtesy of Adobe
Another interesting feature is the Generative Extend tool, which intelligently adds frames to existing shots. This is particularly useful when a clip is shorter than required for the narrative flow.
In practice, these new tools could dramatically streamline how video professionals work. Adobe is also exploring the integration of video generation models from OpenAI and Runway to generate B-roll footage directly within Premiere Pro. As for Generative Extend, it could be used in conjunction with Pika Labs to add extra seconds to the end of a clip.
Image courtesy of Adobe
Addressing the importance of transparency in today’s digital age, Premiere Pro will also incorporate Content Credentials, a digital “nutrition label” designed to inform viewers if AI was used in the creation of media and which models were employed.
[via Adobe, images courtesy]
