Apple Researchers Develop Open-Source AI Model for Image Editing
Apple researchers have recently unveiled a groundbreaking open-source AI model named “MGIE” that has the ability to edit images based on natural language instructions. This latest development showcases the company’s continued commitment to advancing artificial intelligence (AI) and its applications in various fields.
The main feature of MGIE is its utilization of multimodal large language models (MLLMs) to interpret user requests and perform pixel-level manipulations on images. This means that users can simply provide verbal instructions on how they would like their images to be edited, and the AI model will carry out the necessary adjustments accordingly.
With MGIE, users can now make both global and local photo enhancements. Global enhancements include adjusting brightness, contrast, sharpness, and applying artistic effects to images. On the other hand, local editing allows users to modify specific regions or objects in an image, such as changing their shape, size, color, and texture.
Furthermore, MGIE also provides Photoshop-style modifications, including cropping, resizing, rotating, adding filters, changing backgrounds, and blending images. This comprehensive range of editing options gives users the freedom to truly personalize their images according to their preferences.
One noteworthy aspect of MGIE is its incorporation of common sense reasoning. The AI model is able to understand user instructions and make appropriate edits, even going as far as suggesting healthy modifications to certain images. For example, it can add vegetable toppings to a pizza to make it appear healthier.
Apple partnered with researchers from the University of California to develop MGIE, and the open-source model is now available on GitHub, allowing developers and AI enthusiasts to explore its capabilities.
This marks Apple’s second major breakthrough in AI research in recent months. Previously, the company successfully developed a flash memory utilization technique for deploying large language models on Apple devices. These achievements demonstrate Apple’s commitment to pushing the boundaries of AI technology and incorporating it into their products.
In fact, Apple has been placing strong emphasis on AI research and development as it works towards introducing generative AI functionality on the iPhone and iPad by late 2024. The upcoming iOS 18 update is also expected to include an enhanced version of Siri with ChatGPT-like capabilities. If this prediction holds true, it could potentially be the biggest software update in iPhone history.
The introduction of MGIE and Apple’s ongoing AI research projects signify the company’s dedication to innovation and the integration of AI into everyday life. As AI technology continues to advance, we can expect exciting developments from Apple that will shape the future of image editing and other AI applications.
“Bacon trailblazer. Certified coffee maven. Zombie lover. Tv specialist. Freelance communicator.”