Speech Recognition Library in Python always returns same string

Question

I'm trying to use Google Speech recognition, My problem is that after saying something into microphone, results are always same.

My Function looks like this

def RecordAudio():
    import speech_recognition as sr

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

    # Speech recognition using Google Speech Recognition
    try:
        print("You said: " + r.recognize_google(audio))
    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))

    return audio

When I'm trying to get data

data = RecordAudio()

Data is always equals to this speech_recognition.AudioData object at 0x111a8b358

But what is strange for me is that, when this code is in one file, without function and without returning result, in that time this works.

For example I've test.py file

import speech_recognition as sr

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

# Speech recognition using Google Speech Recognition
try:
    print("You said: " + r.recognize_google(audio))
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))

This is working but first example not.

Any idea what's happening ?

You are returning the value of `audio`, which is the raw recorded sound. The result of speech recognition comes from `r.recognize_google(audio)`, which you print out but do not save anywhere. You'd need to save that in a local variable so that you can return it later. — jasonharper, Aug 14 '17 at 14:08
Yes you are right, before you answered it, I solved it and post answer for it. Thank you anyway for response. — Tornike Gomareli, Aug 14 '17 at 14:17

score 0 · Accepted Answer · answered Aug 14 '17 at 14:17

I Solved it.

I think that my problem was, that this function

audio = r.listen(source)

returns AudioData and when I was printing it thats why it was speech_recognition.AudioData object at 0x111a8b358

To recognize it we must return

return r.recognize_google(audio)

It will return real string and not AudioData.

If somebody will have same problem,hope it helps.

Speech Recognition Library in Python always returns same string

1 Answers1