Developing a statistical model for walking based on sensor data from a wrist-worn motion sensing device (accelerometer and gyroscope).

Capturing the variability in walking for a single human subject as well as across multiple subjects. Based on this model, the students will seek to classify types of walking and classify the types of users based only on sensor data.

Project requires undertaking a literature survey on Human Activity Recognition (HAR), access publicly available databases of motion sensor data,and develop software tools to process this data and visualize the results.

Rithmio can sponsor additional devices for data collection.

See: Survey paper on HAR

The data set is not attached in the github repository due to size. It can be unpacked at the following link.

https://archive.ics.uci.edu/ml/machine-learning-databases/00240/