Apple is now ready to take on premium editing softwares with its AI model that can edit content in a photo simply by explaining what you want changed in the image. We all know what DALL-E from OpenAI or Gemini can offer with its AI capabilities and Apple is going to bring further challenges and options for the users.
Apple has relied on education universities to build the image editing language model called the MGIE which is MLLM-Guided Image Editing. The MGIE model can be used to edit simple images and even add a different touch to make it look brighter or even add a new shape. Apple has worked with the University of California, Santa Barbara for the model which can be used to crop, resize and add filters to an image just by using text prompts. It is interesting the way the company uses two language models, first to train the AI to understand the prompts and how to interpret them.
Once the learning comes through, the model starts to imagine how the edits will visually look like and it can be a simple aspect of the edit. Apple is definitely working on a broad set of AI tools and we will probably see all of this combined in the near future and packaged as a part of the iOS or iPadOS platform for its devices.
Make no mistake, Apple is no match for the OpenAI’s or the Adobe’s of the world because of its limited dataset for generative AI applications but that is surely going to change sooner rather than later. The biggest differentiator for Apple in the AI world has to be its on-device processing which makes the whole AI game secure and private for the consumers.
It is exciting to see AI shape up in a big way and 2024 is likely to be a pivotal part of the revolution in the industry and Apple will surely be in the mix.
Comments
0 comment