whisk-ml/bike_image_classifier_tensorflow

Classifier to determine if an image is of a Mountain bike or a Road bike. Built with whisk and utilizing Tensorflow

Bike Image Classifier using Tensorflow

This project was created using whisk. whisk creates a logical and flexible project structure for ML with reproducible results and lets you release your model to the world without becoming a software engineer.

Prerequisites

The following is required to run this project:

• Git (configured with a user name and email)
• Python 3.6+
• A Linux-based OS (includes OSX)

Setup

After cloning this repo and cd bike_image_classifier_tensorflow:

1. If you haven't yet installed whisk, run pip install whisk
2. Run whisk setup. The install script creates a venv, installs the Python dependencies specified, and initializes DVC.
3. Activate the venv: source venv/bin/activate
4. If DVC is used, Download the latest data files: dvc pull.

To learn more about whisk, here are a few helpful doc pages:

• A Quick Tour of whisk
• Key Concepts
• Project Structure

Quickstart

After running the setup, you an invoke the model from the command line. We have a few bike examples hosted for testing:

• Mountain Bike Example
• Road Bike Example

$ bike_image_classifier predict [\"https://whisk-examples.s3.amazonaws.com/bike-images/mountain_bike.jpg\"]

Mountain: 0.9900122, Road: 0.009987746

$ bike_image_classifier predict [\"https://whisk-examples.s3.amazonaws.com/bike-images/road_bike.jpg\"]

Mountain: 0.12031126, Road: 0.87968874

---

Project built with the whisk ML project framework based on the cookiecutter data science project template. #cookiecutterdatascience