This repo is a Android Studio Project with Kotlin as Front-End and Python as Back-End using Chaquopy 🐼
This project was completed in collaboration with the Center of Excellence Artificial Intelligence and Robotics
The Facial Recognition used in this app runs entirely using the face_recognition library in Python. The model used for facial recognition has a 99.38% accuracy on the Labelled Faces in the Wild dataset. Python is linked to Android Studio using the open source Chaquopy plugin. All facial recognition data is stored on a Firebase database.
During the facial recognition stage, the front camera of the Android device is used to capture the image. Then the image is passed as a raw byte stream to the Python backend to extract the facial features. This is then compared against the feature encoding stored in the Firebase database to validate the face.
-
Open Weather Map API: This is used to get weather updates on the Dashboard. It calls for update every hour and also if the distance travelled exceeds 50km.
-
Google Maps API: This is used to display the map on the Dashboard. It is also used to calculate the speed of the vehicle.
-
YouTube Music API: This is used to make the Music Player on the Dashboard. It uses the ytmusicapi and pafy library in Python.
-
Photon OpenStreetMaps API: This is used to get the search results for places in navigation.
The app could be connected to an Arduino on the vehicle and when logged in, the app sends a message to the Arduino via Bluetooth to switch on the relay and start the engine.
For doubts email me at: atinsaki@gmail.com