This repository contains the android application, which is able to detect poses using the MLkit pose detection API, and classify them using a keras sequential model converted to tensorflow lite. The code for the keras/tensorflow lite model can be found in Classifier.py. To run the application, do:
- clone repository
- open app folder using Android Studio
- run application on connected device, or on an AVD (Android virtual device)
Note: to keep the cpu usage as small as possible, no graphic overlay is used. The classification of the pose can be found in the console.