I have already a basic design for speech recognition on paper.

With my increasing understanding what is needed to recognize object groups as well as pure sound input I want to revise the design before I put it here.

Sorry, you have to wait a bit !