Given an image of a dog, algorithm will identify an estimate of the canine’s breed out of total 133 dog's breed.
Keras, Tensorflow, Jupyter notebook, Python 2.7
Download data from https://s3-us-west-1.amazonaws.com/udacity-aind/dog-project/dogImages.zip and make a folder called dogImages inside main project and extract files from given link iside this folder itself.
Make a folder called 'bottleneck_features' inside main project. Download data from https://s3-us-west-1.amazonaws.com/udacity-aind/dog-project/DogVGG16Data.npz and keep it inside 'bottleneck_features' folder.
If you want to make bottleneck features see, 'Making Bottleneck.ipynb' . Since, making bottleneck features is a computationally heavy task it is recommended to use bootleneck features directly.
Run Dog breed identification.ipynb for checking results. You can use pretrained weights saved inside saved_models directory and model_architecture_vgg_self_made.json to directly run model with weights without training.