Speech Recognition - Speech to Text

pip install PyAudio

Note: If you have Python 3.7 see below in Files Attached section for the unofficial WHL file for the PyAudio for Python 3.7

(Obtained from this website:

https://www.lfd.uci.edu/~gohlke/pythonlibs/

Alternatively:  https://github.com/intxcc/pyaudio_portaudio/releases

Move your WHL files into a suitable folder in your Python libraries

such as in folder below pip.exe

create a folder: c:\....\python38\scripts\wheels

 

to install a whl file:

pip install wheels\some-whl-file.whl

(If not, research offline modes for SpeechRecognition)

How Speech Recognition works...

1. Capture sound from microphone.

a) You say something into your microphone - this is captured as a sound file by the Speech Recognition object/class Recognizer() via the listen() method from your microphone as the source, and stored in a python variable audio 

2. The audio file is translated into ASCII text through the Google API..

a) The Recognizer object (r) has a method called "recognize_google"
b) You pass the audio variable/data to this method
c) Google on the internet converts the spoken sound to written text. (essentially using the Google Translate tool API) and returns it from the "recognize_google" call.

Sample Code:

#!/usr/bin/env python3
# Requires PyAudio and PySpeech.

import speech_recognition as sr

# Record Audio
r = sr.Recognizer()
with sr.Microphone() as source:
    print("Say something!")
    audio = r.listen(source)

# Speech recognition using Google Speech Recognition
try:
    # for testing purposes, we're just using the default API key
    # to use another API key, use `r.recognize_google(audio, key="GOOGLE_SPEECH_RECOGNITION_API_KEY")`
    # instead of `r.recognize_google(audio)`
    speech = r.recognize_google(audio)
    print("You said: " + speech)
except sr.UnknownValueError:
    print("Google Speech Recognition could not understand audio")
except sr.RequestError as e:
    print("Could not request results from Google Speech Recognition service; {0}".format(e))

Using Web Based Speech Recognition