Apple has recently unveiled an open-source AI image editing tool called MGIE (MLLM-Guided Image Editing). This tool leverages multimodal LLMs (Language Models for Image Editing) to interpret text-based commands and translate them into pixel-level edits, enabling users to perform a wide range of image editing tasks.
The tool can handle various editing aspects, such as Photoshop-style modification, global photo optimization, and local editing. The tool can crop, resize, rotate, flip, add filters, change backgrounds, add or remove objects, blend images, and apply advanced adjustments to overall photo quality, brightness, contrast, sharpness, and color balance.
Additionally, it can apply artistic effects like sketching, painting, and cartooning. The tool’s advanced capabilities also allow users to edit specific regions or objects within an image, such as a person’s face, and modify their attributes.
“Instead of brief but ambiguous guidance, MGIE derives explicit visual-aware intention and leads to reasonable image editing.” the research paper reads.
“We conduct extensive studies from various editing aspects and demonstrate that our MGIE effectively improves performance while maintaining competitive efficiency. We also believe the MLLM-guided framework can contribute to future vision-and-language research,” the research paper added.
Earlier this month, Apple CEO Tim Cook teased several AI-related announcements for later this year during an earnings call, telling investors he was, “excited to share the details of our ongoing work in that space later this year.”
This AI model represents a significant advancement in the field of image editing, offering a powerful and versatile tool for both professional and casual users.
Currently, the tool is not widely available to the public as an official development from Apple, but users can access it through GitHub for technical exploration or try its web demo on Hugging Face.