mrankitvish / sd-style-photo

SD Style Photo is a Stable Diffusion extention for changing styles of images.

Text-to-Image and Image-to-Image with ControlNet Generation for Stable Diffusion

This project uses the Stable Diffusion API to generate images from text prompts and apply styles to images. It provides a user interface with two tabs: "Create Image" and "Style Image".

Create Image

In the "Create Image" tab, you can input a positive prompt and a negative prompt to generate an image. You can also adjust the number of steps for the image generation process. After clicking the "Generate" button, the generated image will be displayed.

Style Image

In the "Style Image" tab, you can upload an image and apply a style to it. The available styles are fetched from the Stable Diffusion API. You can also adjust the denoising strength and the number of steps for the image-to-image process. Additionally, you can select a ControlNet module and model to further refine the image-to-image process. After clicking the "Convert" button, the styled image will be displayed.

Dependencies

This project requires the following dependencies:

• requests
• gradio
• python-dotenv

To install dependencies:

pip install -r requirements.txt

Usage

To run the project, simply execute the script using a Python interpreter:

python style-photo.py

This will launch the Gradio interface in your web browser. You can then interact with the interface to generate images from text prompts and apply styles to images.

Notes

• The url variable in the script should be set to the address of your Stable Diffusion API instance.
• The controlnetModule and controlnetModel variables in the "Style Image" tab can be adjusted to use different ControlNet modules and models.
• The denoising_strength and img2imgSteps variables in the "Style Image" tab can be adjusted to control the denoising strength and number of steps for the image-to-image process.