Sketching has long been an activity done by people since their childhood.
All it takes, is a piece of paper (or practically anything for that matter), a writing tool, and any form of imagination. It's literally safe to say that everyone can sketch, but not that many people can draw.
This time, Stability AI, the company that created the popular image-generating model Stable Diffusion, launches another AI that can draw better than most people.
But not just drawing from pure imagination because this AI here can draw any image from just sketch input.
The sketch-to-image service, 'Stable Doodle', leverages a Stable Diffusion model to analyze the outline of a sketch and generate a "visually pleasing" artistic rendition of it.
According to Stability AI in a blog post:
"With Stable Doodle, anyone with basic drawing skills and online access can generate high-quality original images in seconds."
Initially only available through ClipDrop, a platform Stability AI acquired in March through its purchase of Init ML, an AI startup founded by ex-Googlers, Stable Doodle takes AI imagination to the next level.
While there’s plenty of sketch-to-image AI tools out there, including open source projects and ad-supported apps, Stable Doodle is unique.
This is because it allows users to fine-tune the process, by giving them a more “precise” control over the image generation, Stability AI explained.
The Stable Diffusion model it uses is called the 'Stable Diffusion XL'.
The model is paired with a "conditional control solution" developed by one of Tencent’s research and development divisions, the Applied Research Center (ARC).
Called the 'T2I-Adapter', it allows both Stable Diffusion XL to accept sketches as input and guides the model to enable better fine-tuning of the output artwork.
“T2I-Adapter enable[s] Stable Doodle to understand the outlines of sketches and generate images based on prompts combined with the outlines defined by the model,” Stability AI explains in the blog post.
In addition to a sketch, Stable Doodle also accepts a prompt to guide the image generation process.
Users are presented with a field box, where they can describe what the AI should understand.
However, there’s a limit to the customization, though.
At least during launch, Stable Doodle only supports 14 styles of art.
Stability AI envisions Stable Doodle to serve as a tool for designers, illustrators and other professionals alike.
With it, they can "free up valuable time" and "maximize efficiency" of their work.
However, the company warns users, that the quality of the images generated by Stable Doodle may not be as high as some people require in their professional work.
"Ideas drawn as sketches can be immediately implemented into works to create designs for clients, material for presentation decks and websites or even create logos," the company said.
"Moving forward, Stable Dooble will enable users to import a sketch. Further, we will include use cases for specific verticals, including real estate applications, for example."
With tools like Stable Doodle, Stability AI is looking forward towards more success following the immensely popular Stable Diffusion.