An Android Application for Fall Detection and Alerts Generation

This video demonstrates an Android application that monitors device sensors in background to calculate acceleration and pitch, uses Weka Multilayer Perceptron classifier for real-time pattern recognition, and generates alerts on detecting a fall. It tackles the problem of false positives using a combination of speech recognition and a with-audio countdown.

I am a Master’s student at NC State University, and this application is developed towards my Master’s thesis.

You May Also Like