Stability AI presents Stable Doodle, a groundbreaking sketch-to-image tool that seamlessly converts simple drawings into dynamic images. It offers limitless creative possibilities for professionals and enthusiasts alike.
I created a few sketches, and with the help of Stable Doodle, they were transformed into stunning images. Here are some: (note: I'm not an artist)
Breathing life into drawings has never been easier after the AI revolution. Stable Doodle, developed by Clipdrop in collaboration with Stability AI, has the potential to revolutionize various industries. It can revolutionize education, design, fashion, and the arts. Users can experience Stable Doodle for free on the Clipdrop by Stability AI website, featuring the latest Stable diffusion model, SDXL 0.9.
Designed for both seasoned experts and beginners, Stable Doodle empowers individuals with basic drawing skills. All that is required is internet access to generate high-quality, original images within seconds.
This user-friendly approach not only saves valuable time but also maximizes efficiency for designers, illustrators, and other professionals. Ideas initially sketched on paper can now be swiftly transformed into designs for clients, assets for presentations and websites, or even stunning logos. The possibilities are limitless, and the process is incredibly enjoyable!
The Technology Behind Stable Doodle
Stable Doodle merges Stability AI's advanced image-generating technology, the Stable Diffusion XL, with the powerful T2I-Adapter. Developed by Tencent ARC, the T2I-Adapter serves as a versatile condition control solution, granting precise control over AI image generation.
By introducing trainable parameters to existing large diffusion models, the T2I-Adapter allows for the inclusion of additional input conditions, such as sketches, segmentation maps, or key poses. This framework supports multiple models for input guidance simultaneously, providing enhanced control throughout the generation process. In the case of Stable Doodle, the T2I-Adapter supplements the pre-trained text-to-image model (SDXL). It enables it to interpret sketch outlines and generate images based on prompts combined with the outlined instructions defined by the model.
With approximately 77M parameters, the T2I-Adapter network offers supplementary guidance to pre-trained text-to-image models (SDXL). It does that while preserving the integrity of the original large-scale models.
Availability and Artistic Customization
Stable Doodle is accessible through the Clipdrop by Stability AI website and app (available on iOS and Google Play). Users can begin experimenting with the tool without the need for a login, subject to daily limits. The intuitive user interface ensures a seamless experience: simply create a basic sketch using a mouse, select an art style, and click ‘generate'. It's that simple!
Moreover, Stable Doodle allows for artistic customization, offering a diverse range of 14 styles through Stable Diffusion XL. Styles vary from realistic (photography) to cinematic and creative (fantasy art and origami), catering to different artistic preferences and inspiring endless creativity.