Smoking DetectorAndroid | Machine Learning | Gesture Recognition |
Quantified Self It is an Android application for both a smartphone and a smartwatch which detects smoking gestures and creates reports for the user.
This project was developed during UBISS 2015 in Workshop A: Sensor-Based Intelligent Mobile Interfaces. It was developed in a group of 3 students and the main focus was using Machine Learning on a smartphone. |
|
We won the Distinguished Project Award on our workshop.
The motivation behind it was to raise smoking awareness and to help research done in smoking behavior and addiction.
A Random Forest runs in the smartwatch to detect puffing gestures every 30 seconds. Then the smartwatch sends data to the phone which creates visualizations of the data.
Next steps:
The motivation behind it was to raise smoking awareness and to help research done in smoking behavior and addiction.
A Random Forest runs in the smartwatch to detect puffing gestures every 30 seconds. Then the smartwatch sends data to the phone which creates visualizations of the data.
Next steps:
- Add the data to the charts.
- Add a daily, weekly or monthly limit capability.
- Add alerts.
Click to set custom HTML