Bot voice call?

Question

Neural networks

I would like to know if there are any materials on setting up a call to your bot with a voice command in the application? (we are talking about methods, and general ideas about the implementation)
Of the materials found, I only met habr with a Yandex article about Alice

In our case, this is solved using a special neural network, which is trained exclusively to recognize key phrases (“Hi, Alice”, “Listen, Yandex” and some others). Support for a limited number of such phrases allows you to perform this work locally and without contacting the server.
If the network is learning to understand only a few phrases, you might think that doing this is quite simple and fast. But no. People pronounce phrases far from ideal conditions, but surrounded by completely unpredictable noise. Yes, everyone has different voices. Therefore, to understand just one phrase, thousands of training records are needed.
Even a small local neural network consumes resources: you can’t just take and start processing the entire stream from the microphone. Therefore, a less heavy-weight algorithm is used on the front line, which cheaply and quickly recognizes the event "speech began". It is he who turns on the neural network engine for recognizing key phrases, which in turn launches the hardest part - speech recognition.

For Google, I could not specifically formulate a request - maybe only I was a Chukchi and could not)

Answer 1

You need to wake word detection - activating the bot with a word/phrase ("Ok Google!"). Something like this .
For voice commands (Speech-to-Intent engine) you can look at this