An issue faced during the automatic seal counting project was the lack of data. More specifically the lack of 150x150 sub-images contains seals. To solve this problem, I have decided to generate additional data utilizing a DCGAN.
NOTE: This repository is missing the training data used to train the DCGAN
- This project utilizes Python 3.10
- Utilizing CUDA is highly recommended (More information can be found here: https://pytorch.org/get-started/locally/)
Install dependencies:
pip install -r requirements.txt