There is StableDiffusion integration plugin in illustration tool called Krita, and it seems image generation from scribble is possible so I try it.
Setting Procedure of Krita and Plugin
1. Installation of Krita
Binary for installation seems prepared for each OS.
Krita はプロフェッショナルな機能を持つ、自由かつ無料でオープンソースのペイントプログラムです。すべての人が入手できるアートツールが存在することを願うアーティスト達によって開発されています。
2. Download Extension for StableDiffusion
First, download krita-ai-diffusion from below. I think latest is fine if there is no particular reason.
Streamlined interface for generating images with AI in Krita. Inpaint and outpaint with optional text prompt, no tweaking required. - Acly/krita-ai-diffusion
3. Install Extension for StableDiffusion
Import downloaded zip from UI of Krita. Restart Krita once imported.
- Tools > Script > Import Python Plugin from file
4. Open illustration screen for now.
Since setting screen cannot be opened unless it is illustration screen, open “New Image” screen.
5. Display panel of Stable Diffusion
Since panel for AI generation is not displayed by default, display it.
It is OK if panel like image below is displayed at bottom right. Click Configure and open setting screen.
6. Install server and model for AI generation
Click Local Managed Server, and click Install button displayed
Simultaneously SD1.5 and CLIP model are also installed so it takes some time.
Try generating actually
Try real time generation (Use Flat2D-Animerge installed by default for model for now).
Display real time generation function from Live button of image below, and start Live Preview by clicking Play (?) button.
Generation Result
Since I am unfamiliar with this kind of paint tool (especially around layer), I compromised, but I think I could generate reasonably as I thought.
