Apple’s AI Image Editor Challenges OpenAI and Google by Accomplishing Your Desires
Apple is now ready to take on premium editing software with its AI model that can edit the content of a photo simply by explaining what you want to change in the image. We all know what OpenAI or Gemini’s DALL-E can offer with its AI capabilities, and Apple brings more challenges and options to users.
Apple has relied on educational universities to build an image processing language model called MGIE, which is MLLM-driven image processing. The MGIE template can be used to edit simple images and even add a different look to make it look brighter or even add a new shape. Apple has worked with the University of California, Santa Barbara to create a template that can be used to crop, resize, and add filters to an image using text prompts. It’s interesting how the company uses two language models, first to train the AI to understand and interpret the prompts.
Once the learning is done, the model begins to imagine what the edits will look like visually, and that can be a simple part of the edit. Apple is definitely working on a wide range of AI tools, and we’ll probably see all of this combined as part of the iOS or iPadOS platform for its devices in the near future.
Make no mistake, Apple is no match for OpenAI or Adobe as it has a limited database for generative AI applications, but that will surely change sooner or later. Apple’s biggest differentiator in the AI world has to be its on-device processing that makes the entire AI game safe and private for consumers.
It’s exciting to see AI develop in a big way, and 2024 is likely to be a key part of the revolution in the industry, and Apple is sure to be a part of it.
