The data for this project is sourced from the Stanford Dogs Dataset.
Python Notebooks:
- exploratory_and_data_preparation.ipynb : EDA and image and datasets preparation
- transfer_learning.ipynb : Xception transfer learning and training
- model_architecture_development.ipynb : data augmentation and CNN architecture development with keras and mlflow
- nn_evaluation.ipynb : benchmark between transfer learning model and CNN specific architecture model
Script for Image Classification:
- run_classification.py : image classification script using transfer learning model
- etat_de_lart.pdf : state of the art of image neural network overview
- pesentation.pdf : presentation slides