This project focuses on implementing Generative Adversarial Networks (GANs) and Variational Autoencoders (VAEs) to generate meander channel images. The dataset consists of 50,000 segmented meander channel images categorized into channel and non-channel regions.
- The main project is located in the
ltrace
folder. - Place the dataset in the
ltrace/datasets/
folder. - The
gan_models
notebook contains the models for GANs. - The
vaes_models
notebook contains the models for VAEs. - Trained models are stored in the
ltrace/trained_models/
folder.
python>=3.10
poetry
Poetry is a modern dependency manager for Python.
- Clone this repository using Git:
git clone git@github.com:AnthonyAposta/GenerativeModels.git
- Install dependencies using poetry:
poetry install
- Activate the Python environment shell:
poetry shell
- Launch JupyterLab:
python -m jupyterlab