For the way to interact, we don’t have a definitive solution. It seems that on the market there are facial recognition solutions that automatically pick up when the person is in front of the tablet, but it’s really expensive (It’s a french product, you can look at https://linote.fr/).

For a start, I’m just thinking about a big button that blinks, and it picks up when you press it. For the people we’re helping, it would be a step forward.

To call someone, it could be a voice command. You press the button, you say the name of the person, and the voice recognition will call the concerned person.

Thanks for your link, I’ll look at it!

https://pusher.com/tutorials/webrtc-video-call-app-nodejs

how do you think the users might interact w the mirror to invoke this function. all of the ways i can imagine take a lot of cognitive skill.

speaking
waving/gestures
typing

even inbound/receiving a call demands significant skill