The long-awaited moment has arrived for video editors and motion designers: the application of Generative Fill to videos. This new technology, previously used in Adobe Photoshop, will now be integrated into After Effects, harnessing artificial intelligence to generate missing or replaceable parts of images based on textual instructions, revolutionizing the workflow and greatly simplifying the work of creatives.

In recent months, there has been a growing implementation of artificial intelligence-based tools in video editing. This trend involves not only Adobe Premiere Pro but also competitors like Davinci Resolve.

At the Adobe MAX 2023 conference, Adobe showcases some of its most innovative technologies. One that particularly caught our attention is undoubtedly Project Fast Fill, a new experimental technology that introduces, for the first time, the capabilities of Firefly, the suite of generative artificial intelligence tools, to the world of AI video.

What is Project Fast Fill?

Project Fast Fill is the AI generative fill for video. It allows users to replace a selected portion of a video frame by inputting a text message to generate new content, which is then automatically propagated to the remaining frames by tracking and animating them in the timeline.

Adobe may have a significant advantage in this field due to its work in creating the Firefly models.

According to Adobe, the technology provides ‘a first look at what human-guided generative artificial intelligence could enable within Adobe’s video editing tools, including Premiere Pro and After Effects.

Generative Fill to After Effects

Project Fast Fill allows editors to remove objects from a video or change backgrounds as if they were working with a still image, all with a simple textual command. The video editing then automatically propagates to the rest of the video scene. Adobe claims that this new tool also works in highly complex scenes with changing lighting conditions.

For more info check here: “Read Adobe’s announcement of Project Fast Fill

Article Comments