Apple introduces an AI image manipulator, and we have the screenshots

Apple recently entered the field of generative AI with the introduction of MGIE, a model capable of manipulating images based on natural language prompts. Developed in collaboration with researchers from UC Santa Barbara, MGIE offers users a novel way to edit photos by simply describing their desired changes in plain text.

Here are the screenshots to understand the functionality.

This model exhibits versatility in handling various editing tasks, ranging from basic operations like cropping and resizing to more intricate modifications such as altering object shapes or adjusting brightness. It combines two distinct AI functionalities: interpreting and translating user prompts into pixel-level image manipulations.

Key characteristics of MGIE would be:

• Edits are initiated through text descriptions, eliminating the need for specialized software knowledge.
• Handles both simple and complex tasks, catering to varying user requirements.
• Freely accessible on GitHub, encouraging exploration and experimentation within the developer community.

While Apple’s plans for MGIE remain undisclosed, its release reflects its growing interest in integrating AI features into its offerings. This move aligns with a broader trend in the tech industry, with established players like Microsoft, Meta, and Google already actively involved in generative AI development.

It’s important to note that MGIE is currently in its early stages of development, and its long-term impact remains to be seen. However, its introduction signifies a significant step for Apple in AI, lagging behind its competitors. 

More here.