[sansnomme]

The short answer: nope. It's a bit like speech recognition and synthesis. The open source stuff never matches up to commercial grades services by AWS and Google.