Back to all articles
•Technology

Adobe's Project Indigo Enhances iOS Camera App with AI Features

View original source

Adobe has announced new AI-powered features for its experimental iOS camera app, Project Indigo, which was launched last year. This update aims to enhance mobile photography through advanced AI capabilities.

  • Key Features:

    • Use of LLMs (large language models) to offer photo critiques and editing suggestions.
    • Introduction of tools for advanced object removal, depth of field generation, and adding various styles to photos.

  • Leadership: Marc Levoy, the head of Adobe's project, who previously worked on Pixel's camera technology, emphasizes the shift from prompt-based editing to deterministic outputs, providing users with straightforward editing options.

  • Comparisons:

    • Unlike Google's Pixel phone camera features, Project Indigo provides more detailed and descriptive feedback, potentially helping users improve their photography skills.

  • Details of AI Features:

    • Photo Critique and Suggestions: The app analyzes and gives professional opinions on framing, lighting, colors, and emotional impact.
    • Edit Suggestions: Offers tips for reshooting to improve framing, exposure, and object placement.

  • Object Removal: While traditional apps require manual selection, Indigo allows toggling for preset items like people, trash, and vehicles to be removed seamlessly.

  • Additional Features:

    • Style transfer allows photos to mimic artistic styles such as watercolor or ink, though the feature is not novel.
    • The AI uses Google's Gemini-based Nano Banana and potentially Adobe's Firefly model. These features are still in testing, accessible to a limited group of users.

Despite Levoy's initial criticisms of prompt-based approaches, Indigo includes a feature for descriptive editing inputs, raising questions about its necessity. Overall, Project Indigo aims to enhance user awareness and skill in photography, beyond mere automation.