Apple AI Model Edits Photos Based on Text Prompts

Apple has launched a new AI model, MLLM-Guided Image Editing (MGIE), which allows users to edit photos using text prompts. This model, developed in collaboration with the University of California, Santa Barbara, offers various image editing tasks such as cropping, resizing, rotating, and adjusting brightness, color balance, and contrast. The introduction of MGIE marks a significant advancement in Apple’s AI capabilities.

How MGIE Works

MGIE enables users to perform complex photo edits by simply typing instructions. This AI model can handle tasks similar to those performed by traditional photo editing software, but with the convenience of text-based commands. For example, users can type “increase brightness” or “crop to a square” to make quick adjustments.

Competitive Edge

A recent conference paper highlights MGIE’s performance improvements and efficiency in image editing. The technology is designed to be accessible and intuitive, making it easier for users to achieve professional-quality photo edits without extensive knowledge of photo editing tools. MGIE’s capabilities are available for technical exploration on GitHub and through a web demo on Hugging Face.

Apple’s AI Strategy

Apple’s development of MGIE aligns with its broader strategy to enhance its AI offerings. In 2023, Apple acquired 32 AI startups, more than Google, Meta, and Microsoft. This aggressive acquisition strategy suggests that Apple aims to integrate advanced AI features into its products and catch up with competitors in the AI and generative AI markets.

Future Prospects

While MGIE is not yet widely available to the public, its development signals Apple’s commitment to leveraging AI for improved user experiences. As Apple continues to expand its AI capabilities, users can expect more innovative features in future Apple devices and software.