Adobe Unveils Beta Release of Generative AI Video Tools

Adobe is making significant strides in the generative AI (GAI) landscape with the introduction of its Firefly Video Model, a powerful addition to its suite of applications. During Adobe MAX, the company unveiled several features now available in beta.

One standout feature is Generative Extend for Premiere Pro, which allows editors to seamlessly incorporate generated footage and audio at the start or end of clips. This tool addresses gaps in transitions or edits and helps maintain consistency by correcting eyelines and actions that may change unexpectedly mid-shot. Designed in collaboration with professional video editors, Generative Extend aims to streamline workflows. However, it’s important to note that the tool is primarily intended for minor adjustments and supports a maximum resolution of 1080p, which may not suffice for high-end cinematic projects.

In addition, Adobe has introduced text-to-video and image-to-video functionalities, akin to offerings from OpenAI's Sora and Meta's Movie Gen. These features are currently in beta within the Firefly web app, accessible via a waitlist. Adobe assures users that the Firefly Video Model and its associated features are safe for commercial use, with Content Credentials watermarks applied to all outputs.

Photoshop is also benefitting from enhancements showcased earlier this year, including Generative Fill and Generative Expand powered by the latest Firefly Image Model, which offers image generation at four times the speed of previous versions. The Generate Similar tool allows users to create variations of objects in their images until they find a suitable option. Both features have been released today.

Moreover, Adobe has refined its Remove tool with a new Distraction Removal feature. This functionality enables users to effortlessly erase common elements—such as stray wires or unwanted tourists from photos—with just a single click.

Lastly, the recent launch of Frame.io V4, the most significant update since the platform's inception nine years ago, further enhances collaborative video and photo production. Adobe has revamped its design to streamline workflows and improve the video playback experience. It was also announced that Canon, Nikon, and Leica will now support the Camera to Cloud (C2C) feature, facilitating direct uploads of media to Frame.io, making it accessible for users of major camera brands.

In summary, Adobe's latest innovations in generative AI and collaborative tools are poised to enhance the creative processes of video and photo editors, providing new opportunities for efficiency and creativity.