Meta has introduced Muse Image, the first image-generation model created by Meta Superintelligence Labs (MSL). According to the tech giant’s blog post, the new AI model is integrated into Meta AI and is intended to generate, edit, and change images using natural language prompts, photos, sketches, and annotations.
The release represents Meta's most recent foray into generative AI, introducing sophisticated image creation capabilities to applications like Facebook, Instagram, WhatsApp, and Messenger.
Muse Image can understand complicated instructions, interact with user-uploaded images, and allow users to make changes by drawing on an image and annotating. This advanced process is completely different from traditional image generators that only use text prompts.
Now, users may create and modify photographs without ever leaving Meta's apps since the model is integrated with the AI chatbot.
According to Meta, Muse Image is intended to increase the flexibility and interactivity of AI image generation. Here are some of its notable characteristics:
Create Pictures Based on Text Prompts: Almost any scenario can be described by users using natural language, and Muse Image will produce an image based on their descriptions. It is designed to comprehend intricate requests with numerous details.
Modify Current Images: Users can upload an old photo and ask Meta AI to alter it rather than creating new ones from scratch. For instance, people can add new items, modify the lighting, and change the background or clothes.
Sketch-Based Editing: The model can also alter photographs with basic sketches or annotations. Users can direct the AI to replace, eliminate, or improve specific elements of an image by drawing over them.
AI-Enabled Instagram Stories Effects: More than 30 new effects enabled by Muse Image will be available on Instagram Stories, helping users create amazing visual effects.
Image Creation for WhatsApp Messages: Picture creation in direct conversations on WhatsApp will be integrated using Meta AI. This feature will initially roll out in some countries and then globally.
Meta AI makes using Muse Image simple. Users can:
Open Meta AI in an app that supports Meta, like Instagram or WhatsApp.
Provide a text prompt to describe the image you wish to generate.
Upload a photo if you want to modify an already-existing image.
Use sketches or annotations to draw attention to particular places that need modification.
Examine the produced image and, if necessary, continue honing it with further cues.
Muse Image will be accessible in Facebook and Messenger as Meta increases its support.
Along with Muse Image, Meta offered a glimpse into its new Muse Video AI model that generates video content based on prompts. Even though the functionality is still under development, it makes it clear that Meta is working towards building not only AI-image generation but also AI-movie generation.
Meta is now focusing more on directly competing with other AI companies that offer image creation solutions and integrate those capabilities into applications used by billions of people.
The startup hopes to increase the accessibility of AI-powered image creation for regular users by integrating text prompts, photo editing, and sketch-based controls into Meta AI.