/generative-data-augmentation

Code for Data Mining Project on Using GANs for Dataset Augmentation

Primary LanguageJupyter Notebook

Generative Algorithms for Data Augmentation

Set Up Instructions

  1. After cloning the project, create a folder in this directory called data.
  2. Download the Fruit 360 dataset from Kaggle, unzip the files, and place them under the data directory.
  3. Install the required packages via pip (pip install -r requirements.txt). It is reccomended that you create a virtual environment for this project.
  4. In each experiment directory, run generate_dataset.py to create a sampled dataset to work from.
  5. To train models, run the appropriate training scripts in the directory.
  6. To test models, edit the test script to load the desired weights then run the test script. It will print out a full classification report.