This is a very rough implementation of a camera on a raspberry pi. Currently, it’s a manually run node server that interfaces with a webcam, sends the output to Google’s Vision API for classification, and speaks the results through headphones.

network camera ml integration

Ideally, this will print to a thermal receipt printer, and will include information about wireless networks. The goal is to attempt to recreate what the machine sees, and to help humans understand the networked environment better.