Adobe is significantly enhancing its experimental iOS camera application, Project Indigo, by integrating a suite of new artificial intelligence features. Launched last year with professional controls and multi-frame super-resolution, the app is now set to offer AI-powered critiques of photographs and advanced editing suggestions, potentially transforming how users approach mobile photography.
At the forefront of these new capabilities is an AI photo critique system, designed to analyse images and provide a "professional" opinion on elements such as framing, lighting, colour composition, and emotional impact. Alongside this, the app will offer capture and edit suggestions, guiding users on how to improve a shot by adjusting framing, exposure, or even identifying specific objects to remove from the viewfinder. For existing photos, the AI can also suggest enhancements utilising Adobe Lightroom controls, bridging the gap between casual snaps and more polished images.
Beyond critical analysis, Project Indigo is introducing sophisticated AI-driven editing tools. A notable advancement is the ability to remove objects from images with unprecedented ease. Unlike traditional methods that require users to manually select or draw around items, the new feature provides toggles for common unwanted elements such as people in the background, litter, wires, poles, fences, and vehicles. Users can also describe custom objects for removal, with early tests showing impressive results in seamlessly eliminating items without introducing visual artefacts. Other additions include AI-generated depth of field to simulate blurred backgrounds and a style transfer feature, allowing users to apply artistic tones like watercolour or monochromatic effects to their photos.
Marc Levoy, who leads the Adobe project and previously developed camera technologies for Pixel phones, highlighted that many generative AI tools often rely on complex prompt-based editing, which can be challenging for users to master. Project Indigo's new experimental features, by contrast, are largely button-driven, aiming to provide more predictable and user-friendly outputs. While some prompt-based editing is available for specific workflows, the emphasis is on accessible, deterministic results.
Adobe is leveraging Google's Gemini-based Nano Banana for these AI functionalities, though the company remains open to integrating other models, including its proprietary Adobe Firefly. These features are currently bundled under an 'AI playground' tab and are in an experimental testing phase, accessible only to a select group of users. There is no guarantee that all these features will eventually roll out to a wider audience, but their development signals Adobe's commitment to exploring the practical applications of AI in photography.