dbvb2k/artisticstyletransfer

Artistic Style Transfer with Stable Diffusion

This project provides a web interface for generating images using Stable Diffusion with various artistic styles and loss functions.

Features

• —Multiple pre-trained artistic styles
• —Various loss functions for different effects
• —User-friendly web interface
• —Real-time image generation
• —Customizable parameters (seed, style, loss type)

Setup

1. 1.Install the required dependencies:

pip install -r requirements.txt

1. 1.Make sure you have the style embeddings in the style_embeds directory:

style_embeds/
    ├── midjourney_embeds.bin
    ├── rahkshi-bionicle_embeds.bin
    ├── black-and-white-design_embeds.bin
    ├── cat-toy_embeds.bin
    └── madhubani-art_embeds.bin

1. 1.Run the web application:

python app.py

Usage

1. 1.Open the web interface in your browser (the URL will be displayed when you run the app)
2. 2.Enter a text prompt describing the image you want to generate
3. 3.Select an artistic style from the dropdown menu
4. 4.Choose a loss function type
5. 5.(Optional) Adjust the random seed
6. 6.Click "Generate" and wait for your image

Loss Functions

• —Blue: Standard blue-channel optimization
• —Color: Color-aware optimization
• —EPS: Epsilon prediction optimization
• —TAP: Texture-aware perceptual loss
• —SA: Self-attention guided loss

Artistic Styles

• —Midjourney: Professional, detailed artistic style
• —Black and White Design: Monochromatic artistic designs
• —Cat Toy: Playful and cute style
• —Madhubani Art: Traditional Indian folk art style
• —Rahkshi Bionicle: Mechanical and sci-fi style

Requirements

• —Python 3.8 or higher
• —CUDA-capable GPU (recommended)
• —8GB+ RAM
• —Internet connection for first-time model downloads

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• —Stable Diffusion by CompVis
• —Hugging Face for model hosting
• —Gradio for the web interface