Logo-Image-Generation

Logo-Image-Generation is a project that implements a Generative Adversarial Network (GAN) to create logo images from random inputs. This project showcases the application of GAN techniques in the field of logo design and generation.

Key Features

• GAN-based logo image generation
• PyTorch implementation
• Customizable image size and batch processing
• Visualization of generated logos
• Training progress tracking

Technologies Used

• Python
• PyTorch
• torchvision
• matplotlib
• tqdm
• IPython

Setup Instructions

1. Clone the repository:

git clone https://github.com/ElFilaliHamza/Logo-Image-Generation.git

2. Download the dataset and place it in the ./dataset directory .

3. Run the Jupyter notebook:

jupyter notebook logo_GAN.ipynb

Project Structure

• logo_GAN.ipynb: Main Jupyter notebook containing the GAN implementation
• README.md: This file, providing an overview of the project
• dataset/: Directory for storing the logo dataset.
• generated/: Directory where generated logo samples are saved

Usage

1. Open and run the logo_GAN.ipynb notebook.
2. Adjust hyperparameters as needed (image size, batch size, learning rate, etc.).
3. Execute the cells to train the GAN and generate logo images.
4. Generated images will be saved in the generated/ directory.

Contributing

Contributions to Logo-Image-Generation are welcome! Please feel free to submit a Pull Request.

License

MIT

Contact

EL FILALI Hamza

Project Link: https://github.com/ElFilaliHamza/Logo-Image-Generation