Coinspeaker
Adobe Announces Firefly AI Model Upgrades
Adobe has unveiled three new generative AI models – Adobe Firefly Image 2 Model, Adobe Firefly Vector Model, and Adobe Firefly Design Model. The new models announced during the company’s Adobe Max event, can create high-quality images, editable vector graphics, and customizable design templates.
The Firefly Image 2 Model is an updated version of the Firefly AI image generator used in popular features such as Photoshop’s Generative Fill. According to Adobe, this latest version has been improved to create higher-resolution images than the original with more vivid colors and color contrast. It can generate more realistic images, rendering better quality details such as foliage, skin texture, hair, hands, and facial features.
Image Model 2 also features AI-enabled editing to help creators customize generated images. Mimicking the use of manual camera controls, users will be able to either manually or automatically adjust the depth of field, motion blur, and field of view of a generated image. The upgrade also has a “Prompt Guidance” feature to help users create more effective text prompts.
Further, a “Generative Match” feature has been added. The feature allows creators to add an image that influences the style of their generated images. Users can either upload an image whose style they would like to replicate or choose one from a preselected list. They can also control the degree of resemblance with a slider. The generated images will automatically be tagged with “content credentials” – a digital label that attaches attribution metadata and marks the image as AI-generated.
In an October 10 blog post discussing the new feature, Adobe’s design leader, Scott Belsky noted that “text prompts alone don’t suffice,” hence the need for Generative Match.
“Finding the exact words to describe the style you have in your mind’s eye can be nearly impossible […] With Generative Match, you can include a reference image along with your text prompt. Firefly’s generative AI will produce imagery that combines both your text prompt and your reference image. If your reference image shows a cat in a whimsical cartoon style, the bear in your new image will sport the same look. If the background in your reference image includes lots of purples, browns, and reds, the woods in your new image will, too,” wrote the company.
Belsky also revealed that the company has put in place “new policies and safeguards” to prevent Generative Match from being abused. Users will be prompted to agree to Adobe’s terms of use and confirm they have rights to use the uploaded image. A thumbnail image of the uploaded content will also be stored on Adobe’s servers to provide a level of accountability. The feature will remain in beta while Adobe seeks feedback. During this period, creators are prohibited from using images that use Generative Match for commercial use. Firefly Image 2 is currently available on the web-based Firefly beta and will “‘soon” roll out to Creative Cloud apps.
Also available in the Firefly beta is the new Firefly Vector model for Adobe Illustrator that allows users to create editable vector images with text prompts. Each element of the graphic is automatically split into “logical” groups and layers.
The Firefly Design model – which powers the new text-to-template beta feature in Adobe Express – generates editable templates for print, social posts, online advertising, video and more.
