Adobe has unveiled a groundbreaking generative fill feature called Project Fast Fill at its MAX conference, offering the capability to effortlessly add or remove objects within videos through the magic of AI. This innovation is part of a series of experimental AI features introduced at the event. Project Fast Fill empowers users to replace clothing accessories on individuals in motion or erase tourists from the background of panoramic landscapes.
The functionality of Fast Fill bears similarities to Google’s Magic Editor, allowing for the repositioning of people or objects, but with a focus on video content instead of static images. Project Fast Fill aims to replicate what Adobe’s Project Stardust accomplishes for still photos, enabling transformations like altering colors with a mere text prompt. Adobe attributes these generative AI features to its enhanced Firefly AI models.
Much like other novel features showcased at Adobe’s conference, Project Fast Fill is currently in the experimental stage, actively being developed by the company’s engineers. However, based on the history of Adobe’s past introductions, similar features have eventually made their way to Creative Cloud users. Although Adobe has its sights set on integrating this new tool into Premiere Pro and After Effects, no specific timeline has been provided.
Adobe is also actively advancing its AI-based editing technology for video, audio, and 3D design in various ways. Project Dub Dub Dub, for instance, can translate voices into different languages. Scene Change facilitates the relocation of subjects to a new setting in a different video clip. Res Up employs diffusion techniques to enhance the resolution of low-quality video. Additionally, Project Poseable is capable of positioning 3D objects based on photographs of real people and generating 3D renderings using text inputs.
As AI generative tools for photos continue to become more accessible to users, exemplified by Google Pixel 8’s Magic Editor, we find ourselves pondering the fundamental question of “What is a photo?” With Adobe’s AI-powered Fast Fill for video manipulation, this transformative ability may soon be extended to a broader audience, leading us to contemplate a similar question: “What is a video?”