Adobe is significantly enhancing its experimental iOS camera application, Project Indigo, with a suite of new artificial intelligence-driven functionalities. These innovative tools are designed to provide users with in-depth critiques of their photographs, offering suggestions on aspects like composition, illumination, color balance, and emotional resonance. Additionally, the update brings advanced editing capabilities such as precise object removal, the ability to generate depth of field, and various artistic style transformations, aiming to make sophisticated photo enhancement more accessible and intuitive for mobile photographers.
Project Indigo, initially unveiled last year, previously offered a range of professional controls, multi-frame super-resolution, and diverse capture modes. The latest additions leverage large language models (LLMs) to analyze images and provide actionable feedback, distinguishing themselves from simpler, prompt-based AI tools by generating more deterministic and consistent outcomes. This approach seeks to guide users toward better photographic practices rather than relying solely on abstract textual commands.
Marc Levoy, the lead figure behind Adobe's Project Indigo and previously instrumental in developing the Pixel phone's camera technology, highlighted the distinction between traditional generative AI and the new features. While many generative AI solutions necessitate intricate prompting, Project Indigo's new AI capabilities are integrated as intuitive buttons, producing predictable and desirable results. This design philosophy empowers users to achieve professional-grade edits without needing extensive technical knowledge or trial-and-error with prompts.
Among the standout features are two critical photography assessment tools. The first provides a "professional" evaluation of the photograph, delving into elements such as framing, lighting conditions, color dynamics, and the overall emotional impact conveyed by the image. This critique aims to offer users a deeper understanding of photographic principles. The second feature delivers tailored capture and editing recommendations, suggesting improvements for re-shooting, such as adjusting exposure or altering the composition. For instance, the app might advise removing distracting elements from the frame, transforming a casual snapshot into a more polished image.
Beyond critical analysis, Project Indigo introduces robust object removal functionality. Unlike conventional methods that require users to manually trace objects, this AI-powered tool offers specific toggles to eliminate common distractions like background individuals, refuse bins, overhead wires, fences, and vehicles. Users can also define custom objects for removal. This feature demonstrates remarkable precision, seamlessly integrating background elements after object deletion, yielding results that are largely free of visual artifacts. Moreover, the app can artificially create a depth-of-field effect, simulating a professional-grade blurred background, which adds an artistic touch to photos.
The application also ventures into creative styling with its style transfer feature, enabling users to transform their images into various artistic renditions, including watercolor, pen and ink, monochromatic, or backlit effects. While style transfer is not a novel concept in mobile photography, Project Indigo's advanced models produce refined outputs. These experimental AI capabilities, consolidated under the "AI playground" tab, are currently in a testing phase, accessible to a select group of users. Adobe remains flexible regarding the underlying AI models, currently utilizing Google's Gemini-based Nano Banana but open to incorporating other models, including its proprietary Adobe Firefly. The ultimate goal is to offer features that not only refine images but also educate users on the nuances of photography, moving beyond mere superficial enhancements.
