Skip to content

Jupyter Notebook for Human Activity Recognition (HAR) with Keras and CoreML

Notifications You must be signed in to change notification settings

ni79ls/har-keras-coreml

Repository files navigation

har-keras-coreml

Human Activity Recognition (HAR) with Keras and CoreML

Keras and Apple's CoreML are a very powerful toolset if you want to quickly deploy a neural network on any iOS device. Most other tutorials focus on the popular MNIST data set for image recognition. We will go beyond this widely covered machine learning example. Instead, you will learn how to process time-sliced, multi-dimensional sensor data.

To be more specific, we will train a deep neural network (DNN) to recognize the type of movement (Walking, Running, Jogging, etc.) based on a given set of accelerometer data from a mobile device carried around a person's waist.

We will use the WISDM data set (Activity Prediction) for this tutorial: http://www.cis.fordham.edu/wisdm/dataset.php

The original Medium article can be found here: https://towardsdatascience.com/human-activity-recognition-har-tutorial-with-keras-and-core-ml-part-1-8c05e365dfa0