Generative Algorithms for Data Augmentation

Set Up Instructions

After cloning the project, create a folder in this directory called data.
Download the Fruit 360 dataset from Kaggle, unzip the files, and place them under the data directory.
Install the required packages via pip (pip install -r requirements.txt). It is reccomended that you create a virtual environment for this project.
In each experiment directory, run generate_dataset.py to create a sampled dataset to work from.
To train models, run the appropriate training scripts in the directory.
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.

Name		Name	Last commit message	Last commit date
Latest commit History 34 Commits
fruit_360_experiments		fruit_360_experiments
mnist_experiments		mnist_experiments
weights		weights
.gitignore		.gitignore
README.md		README.md
Untitled Diagram.drawio		Untitled Diagram.drawio
fruit_360_eda.ipynb		fruit_360_eda.ipynb
gan_output.png		gan_output.png
gan_output_mnist_v1_b_32_e_1000.png		gan_output_mnist_v1_b_32_e_1000.png
gan_output_v1.png		gan_output_v1.png
gan_output_v2.png		gan_output_v2.png
gan_output_v3.png		gan_output_v3.png
requirements.txt		requirements.txt
vae1.py		vae1.py
vae_fruits_output.png		vae_fruits_output.png
vae_mnist_output.png		vae_mnist_output.png