SAN: AI Based Visual Self Awareness
Table of Contents
1 Abstract
Two or maybe four cameras are put into each node. The cameras see the environment and in particular the connections to other nodes. Camera information is sent to the main computer e.g. via WiFi.
The network configuration plus camera images is used as training data for machine learning. Thereby the algorithm learns to infer the network configuration from the camera images.
It is no longer needed for the nodes to communicate.