The inspiration behind this idea were the Nest and August home security products. 'Who is it?' does everything they can and more at half the price.
What it does
'Who is it?' detects knocks on your door using an Arduino. The Arduino sends the signal to the Raspberry Pi, which then takes a picture of the person and passes it to the Google Cloud Services. An analysis of the image is conducted to detect the mood and identity of the person. The user can then ask Alexa about the person at the door and Alexa will give advice. Amazon Alexa uses Amazon Web Services and Firebase to talk to the Raspberry Pi.
How I built it
Code! I used Amazon Web Services to configure Alexa to talk to the Raspberry Pi via Firebase. The Raspberry Pi also connects to Google Cloud Services to analyze the person's mood and uses the camera module to take the image. The Arduino is connected to the Pi using Serial.
Challenges I ran into
Connecting Alexa to Firebase was a big effort. Also trying to download OpenCV on Raspberry Pi took a very very very long time (Slow wifi).
Accomplishments that I'm proud of
The image analysis works as predicted and Alexa speaks as expected as well. The project's hardware works together in harmony.
What I learned
Alexa API, Image Analysis through Pi, Lambda (Amazon Web Services)
What's next for Who is it?
Configure more personality traits like sadness, anxiety etc through the image analysis.