(As of November 30, 2016, I'm still writing notes)
For the time being, I tried to make voice recognition from a microphone with Python. I used Azure's Bing Speech API. https://www.microsoft.com/cognitive-services/en-us/speech-api/documentation/overview
When I created an account and clicked on the trial button etc., a Cognitive Services account was created on the Azure screen, and the KEY information required to use the API came out.
So far, it seems that there are still courses to try for free.
When using it in Python this https://github.com/Uberi/speech_recognition
used. pip install SpeechRecognition It is installed with. It seems to be quite sophisticated.
I also needed to install something called pyaudio. This can also be installed from pip. For the time being, when I put SpeechRecognition and run the sample source, I got an error saying that pyaudio is required, so I installed it after seeing it.
The sample code is https://github.com/Uberi/speech_recognition/blob/master/examples/microphone_recognition.py I used it as it is. It works for the time being.