ControlNet Scribble : Turn Your Sketches Into Stunning Images Yeni, May 21, 2024May 21, 2024 You like sketching but maybe need a little lift to turn your sketches into more life-like images ? We got you! ControlNet Scribble allows users to draw or scribble directly on an image to specify the necessary adjustments, edits, or properties for the created output. In this blog post, we will show you how to use ControlNet Scribble. What is ControlNet Scribble? ControlNet scribble is a feature that creates the appearance of hand-drawn sketches with clear lines, making it ideal for styled graphics. Also, ControlNet Scribble transforms image generation with manual annotations, allowing users to direct the process by indicating desired changes. The model interprets user annotations as commands, ensuring that the resulting output is consistent with the user’s vision. What are the ControlNet Scribble preprocessor in DiffusionHub? We have two kinds of Scribble Prepocessor in DiffusionHub. Xdog: Xdog or Extended Difference of Gaussian (xDoG), is an edge detection technique. Basically, by adjusting the xDoG threshold, you have the power to control the level of detail in your artwork. HED HED is a pre-processor that produces outlines that look like they were hand-drawn by a professional artist. Comparatively, ControlNet’s specialists recommend HED for recoloring and restyling photos, giving them a distinct artistic style. How to use ControlNet Scribble First, login into your DiffusionHub account, select Automatic 1111. In the Text2Image tab, scroll down to ControlNet tab. You can either draw a sketch by using a blank canvas or upload an image. To use a blank canvas, go to the white note icon > create new canvas > adjust the pencil However, in this tutorial, we will upload an image we found on the internet. 1.Upload a sketch image. Click ControlNet Enable 2.In the preprocessor tab, you can choose between scribble xdog or HED. In this tutorial let’s choose xdog. On the model controlnet tab, choose control_v11p_sd15_scribble. 3. You can add prompts as you like, in this case we add: a beautiful butterfly, professional majestic oil painting by Ed Blinkley, highly detailed. 4. Make sure to put SD Unet settings to “none” 5. Click Generate. Voila, Here are the result : Linear architectural sketch detached house Wrap up Stable Diffusion is a powerful AI Tool. With new developments happening everyday, it can be daunting to learn and catch up with Stable Diffusion Skill. Finally, DiffusionHub is commited to bring you more easy tutorial like this so we hope by now you can be more confident in exploring this powerful tool. Lastly, See how easy to turn your sketches to images using DiffusionHub? Try it now for 30 minutes FREE! Share on FacebookPost on XFollow usSave Uncategorized Automatic1111ControlNetImage Generation
Uncategorized Stable Cascade February 21, 2024 Introduction of Stable Cascade Stable Diffusion has come a long way to be an open-sourced AI Image generator in the market. We all know that image generated by AI often has flaws, particularly in Stable Diffusion, with its famous deformed human organs with three fingers or misplaced eyes, not-so-perfect added… Read More
Generative AI Introducing HelloWorld: a realistic SDXL base model February 7, 2024February 14, 2024 Table of Contents: HelloWorld Evolution It’s 2024 and AI photography has evolved so much compared to the last year. The new SDXL model HelloWorld marks a groundbreaking transition from the established SD1.5 platform to the futuristic SDXL series. This new SDXL base model series not only symbolizes a departure… Read More
Uncategorized 5 Cool LoRA to Elevate Your AI Images April 17, 2024April 17, 2024 Stable Diffusion community is always quick with new update, quirky checkpoints and new styles of LoRA. CivitAI is home of most Stable Diffusion creators who are always very generous to share their work to the masses. In this blog post, we round up 5 new cool LoRA for you to… Read More