Adobe’s editing applications, such as Photoshop, Lightroom, and Premiere, have been the cornerstone of many creatives’ everyday workflows. Since the rise of generative AI, Adobe has been developing ways to optimize those workflows, and this one will be a huge help for video editors.
At Adobe MAX, the company’s annual creative conference, the company teased Project Fast Fill for video, which will allow users to enter simple text prompts to edit their video.
Also: Adobe unveils three new generative AI models, including the next generation of Firefly
Specifically, users will be able to remove or add an object, or change background elements in seconds with a quick text prompt, tasks that would typically take intense editing to do.
The company shares that Project Fast Fill harnesses its popular Generative Fill to bring generative AI into its video editing applications.
The demos the company presented at the event showcased different potential uses. In the first example, the presenter, Adobe research engineer Gabriel Huang, cleanly removes people from the background of a video, as seen below.
In the second example, Huang adds a tie to a man who is walking towards the camera by simply inputting the text prompt “tie” into Generative Fill after selecting the area of the collar.
In both instances, Generative Fill generated four different video results, which Huang got to choose from to develop the final product.
Also: How Adobe is leveraging generative AI in customer experience upgrades
Even though the edit is only being done on one frame of the video, much like you would with a photo, the tool is automatically applied to all of the frames of the video and automatically adjusts for lighting changes, shadows, the object moving, and more.
Since it was only a sneak peek, there are no details on when Project Fast Fill will be available, with Adobe saying that the sneak peeks are “cutting-edge, experimental technologies that could someday become features in Adobe products.”
