Adobe adds Generative AI to Photoshop

Adobe’s new Generative Fill function lets you create new objects to add to existing image using simple text prompts, such as ‘Add a lighthouse’. (Source: Adobe.)

The new addition will enable Photoshop users to generate objects to add or extend or remove content from images by using simple text prompts. The new Firefly-powered Generative Fill will automatically match the perspective, lighting and style of images to enable users achieve the results they want while minimising tedious tasks. Firefly has been designed to generate images that are safe for commercial use. It was trained on Adobe Stock’s hundreds of millions of professional-grade, licensed, high-resolution images, helping ensure Firefly won’t generate content based on other peoples’ or brands’ intellectual property.

Users will be able to:
– Add new objects to an image after users provide the description of the addition and location of where it is to be added. Ai processing will combine them.
– Remove unwanted objects from a picture.
– Revise objects through changing colours in selected areas or enlarge or reduce the size of objects.
– Create new backgrounds using text prompts.
– Add special effects to text to create new fonts.

The new content is created in generative layers, which can be edited by the user without affecting the original image.  Generative Fill supports Content Credentials authentication, which plays an essential role in ensuring people know whether a piece of content was created by a human, AI-generated or AI-edited.

Photoshop’s Generative Fill feature is available in the desktop beta app today and will be generally available in the second half of 2023. The beta release of Generative Fill runs in the cloud, which requires an internet connection. As a beta offering, new content is generated at a maximum of 1024 x 1024 pixels sp it can be used freely on internet servers. Generative Fill is also available today as a module within the Firefly beta app.