The features provide breakthrough advancements in GenAI image creation from directly within the image editing app
Adobe used its annual Max event in London this morning (Tuesday 23 April) to unveil the latest version of Photoshop (which is currently in beta). New features include what Adobe describes as “breakthrough advancements” in Generative Fill, as well as Generate Image features directly within the image editing app.
The GenAI features are powered by a new version of Adobe Firefly Image 3 Foundation Model, which promises “massive advancements” in quality and control when using GenAI to create images. It brings even more lifelike and realistic imagery, better understanding of prompts, new levels of detail and variety and significant improvements in the images it creates. See the gallery below for examples.
The new features in Photoshop, using the latest incarnation of Firefly, include Reference Image, Generate Image, Generate Similar, Generate Background and Enhance Detail.
The features “empower everyone to generate infinite background variations, create a multitude of design concepts with differentiated colour and style and bring to life detailed scenes from scratch with complete control” from directly within Photoshop, says Adobe.
Ashley Still, senior vice president, Creative Product Group at Adobe, said: “This massive new update to Photoshop takes creativity to unprecedented levels – whether you are a professional or brand new to Photoshop. Through new Firefly-powered features in Photoshop, we are bringing ideation and creation closer together and making editing images both more powerful and approachable so everyone can realise their creative vision.”
Photoshop - New GenAI features overview:
• Reference Image helps creatives achieve the outputs they imagine, using user-selected images as generative inspiration.
• Text to Image with Generate Image enables full text to image capabilities directly within Photoshop for the first time.
• Generate Background replaces and creates backgrounds, making it easier than ever to generate content that seamlessly blends into existing images.
• Generate Similar lets creators iterate with variations they select, to explore creation ideas more deeply with more precise control.
• Enhance Detail fine-tunes images to improve sharpness and clarity.
Adobe has over a decade’s history of AI innovation. Last year, it added Generative Fill and Generative Expand to Photoshop. Since then, users have generated more than 7 billion images with Firefly and, Adobe says, have adopted Generative Fill at ten times the rate of other popular Photoshop features.
Adobe Firefly was trained on licensed content, including Adobe Stock. It was designed to generate content for commercial use that does not infringe on copyright and other intellectual property right such as trademarks and logos.
Photoshop’s latest release is available today in the Photoshop desktop (beta) app and will be generally available later this year.
You can also access Adobe Firefly image-to-text directly at https://www.adobe.com/uk/products/firefly/features/text-to-image.html
MAIN IMAGE: Image created by Adobe Firefly Image 3 Foundation Model, using Generate Image in Photoshop, with the prompts ‘Cat with pearl earings’
No comments yet