How To Use

1. Upload Your Sketch – Start by uploading a grayscale matte painting, sketch, or line-art drawing. This will serve as the foundation for the full-color rendering.
2. Describe the Scene – Input a short description of the subject matter. This helps guide the model to generate a scene that aligns with your vision.
3. Adjust Sketch Strength – Control how closely the AI should follow your sketch. A lower value allows more creative freedom, while a higher value sticks closely to the original outlines.
4. Select Style Enhancements (Optional) – You can add style refinements such as cinematic lighting, lens effects, or artistic direction to fine-tune the final output.
5. Choose a Model – Select the AI model that best suits your needs for image generation.
6. Generate the Image – Click “Invoke” to process the sketch and apply the colorization and rendering.‍
7. Review & Save – View the result in the gallery, make refinements if necessary, and save the final image to your project board.

How It Works

1. Sketch Preprocessing – The uploaded grayscale sketch is processed using edge detection to extract the structural details.
2. ControlNet Integration – A Hard Edge Detection model refines the input to ensure that key lines and structures are preserved.
3. Prompt Construction – The user-provided description is combined with predefined style prompts to create a detailed AI prompt.
4. AI Rendering – The system uses a Stable Diffusion XL model to interpret the sketch, apply textures, and generate a high-quality, full-color image.
5. Denoising & Refinement – The model undergoes several iterations of denoising, adjusting colors and details to match the desired artistic style.‍
6. Final Processing & Output – The completed image is stored in the user’s gallery, where it can be reviewed, further edited, or exported.

‍