Introduction

This project uses Machine Learning to detect if a person has an irregular gait. Our goal is to use these classifiers as an early warning sign for Parkinson's disease.

Data Collection

We gathered data by filming ourselves walking with irregular and regular cadences. We then processed the data by sampling frames of the videos and passing them through OpenPose. We trained both a CNN and a SVM classifier on the dataset and compare accuracy of both.

How to Run the Code

Run the code from run.py