Adobe presents its “revolutionary” AI tool that automatically recognizes objects in photos and edits them
The Project Stardust editor can automatically identify and manipulate individual objects in images (such as removing unnecessary objects from a photo and adding AI-generated content).
In the demo, the tool automatically recognizes objects in the photo, such as a yellow suitcase and its shadow, and highlights them like the Lasso tool in Photoshop. They can be moved, deleted, and so on – as if they were on a separate layer – and the space behind them is automatically filled in to match the rest of the content.
Project Stardust includes something similar to the “contextual taskbar” that debuted in Photoshop earlier this year, which can automatically identify the next steps in the design process and speed up the editing process. In Adobe’s demonstration, when you select people in the background, the panel will offer to “remove distractions.”
The new tool has some of the same generative AI capabilities as Firefly-based Adobe Photoshop. You can select an area in the photo and use text prompts to fill in the space, for example, with flowers.
In another video, the same function is used to replace individual items of clothing on the model by selecting an item (such as a jacket or sneakers) and then describing the new one to be added.
Automated design tools are becoming increasingly common as generative artificial intelligence develops. Canva has similar editing capabilities available to automatically remove or modify objects in images as Google Photo’s Magic Editor tool that comes with Pixel 8 devices.
At the same time, there are few details about Project Stardust, and the mentioned features seem to be “only a small part” of its capabilities – Adobe promises a new mechanism that will “revolutionize the way we interact with the company’s products.” More details are expected at the Adobe Max event, which starts on October 10.