Adobe has introduced a set of advanced AI-powered features to its experimental iOS camera app, Project Indigo, which debuted last year. The app, led by Marc Levoy, brings new tools that use large language models to critique users' photos, offering professional-style feedback on framing, lighting, colors, and emotional impact. Alongside critiques, the app provides capture and editing suggestions to help users improve their shots, including advice on framing and exposure adjustments, as well as tips for removing unwanted objects from the viewfinder.

One of the standout additions is a sophisticated object removal function that goes beyond traditional manual selection techniques found in apps like Apple Photos and Adobe Photoshop. Project Indigo lets users remove various background elements such as people, trash, fences, wires, or vehicles via toggles, and even supports custom object descriptions. This feature produces impressive results, seamlessly erasing unwanted elements like a person or an item without leaving behind visual artifacts.

The update also includes AI-driven depth of field simulation, allowing users to add a realistic blurred background to their photos, and a style transfer capability that applies artistic effects like watercolor or pen and ink to images. While style transfer has existed in other applications and is not entirely novel, Adobe’s implementation uses advanced models that generate more refined outputs. Additionally, users can describe specific edits they want to make, with these instructions processed by AI models like Google’s Gemini-based Nano Banana or Adobe’s own Firefly.

Currently, these AI enhancements are experimental and available only to a select group of users under the app's AI playground. Adobe remains open to evolving the technology and integrating different AI models, but the company has emphasized the value of features that teach photography techniques through AI critique rather than just simplifying photo editing. It remains uncertain if or when these features will be rolled out to a broader audience, but they represent a notable step in combining creative guidance with AI in mobile photography.