The end goal of this project is to classify retail pairs of sneakers and to also design new models of shoes using Deep Learning techniques such as Convolutional Neural Networks and Generative Adversarial Networks.
The data used for training the model has been scraped and pre-processed using the python script scrap_and_process.py
. The in parameters.json
file contains its parameters.
The models are built from the DeepShoes.ipynb
Notebook ran on Kaggle with our scraped dataset uploaded.
The backup/
folder contains the final versions of the built models, uploaded on Kaggle as well. They are used in the DeepShoes-usage.ipynb
Notebook (that can be run here) to generate shoes' images and predict their model.