🖼️ Text-to-Image Generation with Stable Diffusion:

This project demonstrates how to generate images from text prompts using the Stable Diffusion model in a web interface built with Streamlit. The model takes user inputs in the form of text descriptions (prompts) and generates corresponding images using a pre-trained deep learning model.

📝 Features:

Text-to-Image Generation: Converts a user-provided text prompt into an image. Stable Diffusion: Utilizes the StableDiffusionPipeline from Hugging Face's diffusers library. CUDA/CPU Support: Automatically detects if CUDA (GPU) is available and uses it for faster inference; otherwise, it defaults to CPU. User-friendly Interface: Built with Streamlit to allow users to input prompts and download generated images directly from the web interface.

🛠️ Requirements:

Before running the project, ensure that you have the following installed:

Python 3.8+ pip (Python package installer)

📦 Installation:

Clone the repository:

git clone https://github.com/Panchadip-128/Text-to-image-conversion-tool.git cd text-to-image-generation

Install dependencies:

You can install the required dependencies using the provided requirements.txt file:

pip install -r requirements.txt The dependencies include:

torch: PyTorch for deep learning. diffusers: Provides access to Stable Diffusion models. transformers: Hugging Face’s transformer models. accelerate: Optimizes model training and inference performance. streamlit: For the web app interface. Pillow: Image processing library for saving generated images.

🚀 Running the App:

After installing the dependencies, you can start the Streamlit app by running the following command:

streamlit run app.py This will launch a web interface where you can enter text prompts and generate images based on the provided text.

Example Prompts: "A futuristic city skyline at sunset" "A beautiful landscape from mountains during sunrise" "An astronaut floating in space, looking at Earth"

📁 Project Structure:

. ├── app.py # Main Python file for running the Streamlit app

├── requirements.txt #Project-dependencies ├── README.md # Project documentation └── generated_image.png #Example output image

🔧 How It Works:

Loading the Model: The app uses StableDiffusionPipeline from the Hugging Face diffusers library. The model is loaded from the CompVis/stable-diffusion-v1-4 model repository.

Text Input: Users enter a text prompt via the Streamlit interface.

Image Generation: The text prompt is passed to the Stable Diffusion model, which generates an image corresponding to the description.

Output: The generated image is displayed in the Streamlit app, and the user can save the image locally.

🖥️ Example:

Below are the examples of the interface and a generated image:

Prompt: "A futuristic city skyline at sunset" Output:

Prompt: "Formula one cars in racing tracks" Output:

🧪 Technologies Used:

PyTorch: As the underlying framework for model inference. Hugging Face Diffusers: Provides pre-trained models for diffusion-based image generation. Streamlit: Used to create the interactive web interface for the app. Pillow: For saving and processing images. 🐞 Troubleshooting CUDA Error: Ensure that you have a compatible GPU and the necessary CUDA libraries installed if you want to run the model on GPU. Otherwise, the model will run on the CPU by default. Model Download Issues: If you encounter problems downloading the model, ensure that you have an active internet connection. You may also need to authenticate with Hugging Face if required. 💡 Future Enhancements Model Selection: Add options to select between different versions of the Stable Diffusion model. Image Customization: Allow users to specify image size, style, and other customization options. Batch Processing: Generate multiple images at once based on different prompts.