Machine Learning Nanodegree

Deep Learning

Project: Deep Learning Image Classifier

Project for Udacity's Intro to Machine Learning with TensorFlow Nanodegree program. In this project, I developed code for an image classifier built with TensorFlow, then converted it into a command line application.

In order to complete this project, I used the GPU enabled workspaces within the Udacity classroom.

Install

This project requires Python 3.x and the following Python libraries installed:

• NumPy
• matplotlib
• tensorflow
• tensorflow_hub
• tensorflow_datasets
• json
• time
• os

You will also need to have software installed to run and execute an iPython Notebook

I recommend installion Anaconda, a pre-packaged Python distribution that contains all of the necessary libraries and software for this project.

Code

Template code is provided in the Project_Image_Classifier_Project.ipynb notebook file.

Run

In a terminal or command window, navigate to the top-level project directory Deep-Learning-Image-Classifier/ (that contains this README) and run one of the following commands:

ipython notebook Project_Image_Classifier_Project.ipynb

or

jupyter notebook Project_Image_Classifier_Project.ipynb

This will open the iPython Notebook software in your browser.

To use the command line implementation in a terminal or command window, navigate to the top-level project directory Deep-Learning-Image-Classifier/ (that contains this README) and run one of the following commands:

python predict.py path/to/image path/to/model 

Examble:

python predict.py test_images/cautleya_spicata.jpg best_model.h5

There are also some optional parameters:-

• --top_k returns: the top k classes with their probabilties.
• --category_names: a json file that maps each class number with a class name

Examble:

python predict.py test_images/cautleya_spicata.jpg best_model.h5 --top_k 6 --category_names label_map.json

Data

The data for this project is quite large - in fact, it is so large you cannot upload it onto Github. If you would like the data for this project, you will want to download it from the workspace in the classroom.

%pip --no-cache-dir install tensorflow-datasets --user
%pip --no-cache-dir install tfds-nightly --user

Though actually completing the project is likely not possible on your local machine unless you have a GPU. I trained the deep learning classifier using 102 different types of flowers, where there ~20 images per flower to train on. Then I used the trained classifier to see if I can predict the type for new images of the flowers.

Certification