Adobe has announced the release of their new Firefly Video Model.
They claim the new video generator offers unparalleled creative control and is the only generative AI model that is IP-friendly and commercially safe.
The new Firefly Video Model is integrated within existing Adobe Image, Vector, and Design models.
Key updated features for the Firefly Video Model include:
- Generative Extend - a tool that can intelligently expand existing video clips by generating new frames or audio
- Image to Video - a tool that enables the creation of video from still images or illustrations
- Text to Video - a tool where users can generate video simply by typing a description, with heightened control over “camera” properties like position and movement.
The video model is also adept at generating 2D and 3D animation, including mimicking claymation styles. Users can easily adjust between vertical and horizontal aspect ratios as well as having the ability to add keyframes.
It is also able to create realistic footage of natural scenes including landscapes and animals, producing impressive visual effects such as fire and smoke.
Read: What is Google Veo? Inside the AI Video Generator
However, Adobe admits that Firefly is not all the way there yet on creating videos that include humans, but they are actively working to improve this.
This is a distinct disadvantage for Adobe, with models like ByteDance’s OmniHuman specializing in generating realistic videos of human movement and speech based on a single photograph.
What is Firefly Video Model?
Firefly Video Model is a new AI model developed by Adobe that brings generative AI to already popular production tools including Premiere Pro and After Effects.
Users can generate entirely new clips, alter existing videos or expand and add additional video or audio frames from existing video clips.
Firefly video model will , making it easier to add visual effects, create B-roll footage, and generate entire scenes from scratch
