Adobe Premiere Pro is adding generative AI capabilities

An increasing number of software applications are incorporating generative AI features, and Adobe is no exception. Recently, Adobe released a video on its official YouTube channel showcasing new generative AI capabilities that will be integrated into the upcoming version of Premiere Pro. These features are supported by Firefly’s new video model and utilize third-party AI models from Open AI, Runway, and Pika.

The video highlights three generative AI functions available for free: object addition, deletion, or replacement; generative expansion; and text-to-video conversion. It first demonstrates the item replacement feature, where users simply select the item they wish to replace and describe the desired item in a pop-up dialog, and Firefly automatically generates and inserts the described item into the video. Users can also select items they wish to remove from the scene, and the software will automatically delete them and regenerate the corresponding background details, such as filling in a wall after a sign is removed.

The generative expansion function allows the software to automatically fill in blank spaces between two clips. Users select the segment they wish to extend, and activate Firefly, and the software automatically generates a video segment similar to the last shot of the clip. This feature smoothly bridges two segments, saving both shooting time and costs.

The final feature, text-to-video, enables users to generate desired video segments directly from textual descriptions, useful for conceptualizing or creating videos. According to the video, this feature primarily relies on Open AI’s Sora model and Runway AI, allowing users to choose which large model to use, with the possibility of more video generative models being added in the future.

The release date for these features is currently unclear, but it has been suggested that they might first appear in a beta version of Premiere Pro. Those interested should stay tuned for updates.