Images are obviously not included within this repository. Check this : https://www.kaggle.com/c/dogs-vs-cats/data
You will have to include the unzipped /train folder in the project's root.
- Launch preProcessIntoStandardDirectories to create a test folder.
- Launch createFinalDogs&CatsDatasets to separate cats and dogs images into differents folders.
Once you have all necessary images and folders, you can :
- generate and save a one block model (with convolutions) using baseCnnModel.py
- generate and save a three blocks model (with convolutions) using threeBlocksVggModel.py
- generate and save a model with dropout regularization using preProcessIntoStandardDirectories.py
- generate and save a model trained to recognize multiple species (from finalize_dogs_vs_cats folder) using multipleSpeciesModel.py