Door Lock Detector using Edge Impulse - TinyML

Door Lock Detector Using Edge Impulse Software

This TinyML model analyse the sound near to the door and detects if the door is not locked properly.

What is the impact of this prototype?

In western countries, 30% of burglaries are due to not locking the door properly.

Door Lock Detector using Edge Impulse :

I have collected the audio datasets of Idle, Door locking, Door not Locked in the real time using Arduino Portenta H7. I have trained the model in Edge Impulse and downloaded the firmware and flashed to Portenta H7.

The Door Locking detection involves:

• Door Lock Data Acquisition
• Model Training in Edge Impulse
• Classification
• Deployment

Configuration

After firmware installation, In command window, type the daemon command.

edge-impulse-daemon

Once the device is connected, go to -> Data acquisition section and collect the Door Lock scenario data.

Ensure the data is collected in less noisy environment to capture only the Door Lock detection data.

In a create Impulse section, configure the keras, Time series (window size).

Data is collected in this ratio.

The datasets are classified as :

1. Idle

2. Door Locked

3.Door Not Locked

Neural Network Architecture:

In Neural Network, configure the neural network layer as mentioned below.

Model testing Result:

How to flash in hardware?

After successfully verified the trained model, deploy back to the Portenta H7.

Go to deployment section and select the board.

https://github.com/Manivannan-maker/DoorLoclDetector