Question 103508 in Azure Question posted in
The official documentation to get you started can be found here.

How to save Microsoft Azure speech to text output using continuous recognition to a file using Python?

I’m using Speech to Text on a *.wav file from within the Azure machine learning studio. Since the wav file contains a couple of minutes speech I’m using continuous recognition. I do get the STT working. However, I only get the result on the screen and don’t manage to save it to a file. The file is created, but it is empty.

Can you see what I’m missing here in the recognized_cb function ? My code is written in Python.

 def stt_run4(wav_file_path, taal, key, regio, outputfile):
      
    speech_config = speechsdk.SpeechConfig(subscription=key, region=regio)
    speech_config.speech_recognition_language=taal
    audio_config = speechsdk.audio.AudioConfig(filename=wav_file_path)

    speech_recognizer = speechsdk.SpeechRecognizer(speech_config=speech_config, audio_config=audio_config)
    
    done = False


    # Set up the output file for the transcript
    output_file = open(outputfile, "w")

    def stop_cb(evt):
        """callback that signals to stop continuous recognition upon receiving an event `evt`"""
        print('CLOSING on {}'.format(evt))
            # Close the output file and stop the continuous recognition session
        output_file.close()
        speech_recognizer.stop_continuous_recognition()
        print("Transcript saved in file:", outputfile)
        nonlocal done
        done = True
        
    def recognized_cb(evt : speechsdk.SpeechRecognitionEventArgs) :
        if speechsdk.ResultReason.RecognizingSpeech == evt.result.reason and len(evt.result.text) > 0 :
            print('RECOGNIZED:', evt.result.text)
            output_file.write(evt.result.text)
            output_file.flush()
            


    # Connect callbacks to the events fired by the speech recognizer
    speech_recognizer.recognizing.connect(lambda evt: print('RECOGNIZING: {}'.format(evt)))
    speech_recognizer.recognized.connect(recognized_cb)
    speech_recognizer.session_started.connect(lambda evt: print('SESSION STARTED: {}'.format(evt)))
    speech_recognizer.session_stopped.connect(lambda evt: print('SESSION STOPPED {}'.format(evt)))
    speech_recognizer.canceled.connect(lambda evt: print('CANCELED {}'.format(evt)))
    # stop continuous recognition on either session stopped or canceled events
    speech_recognizer.session_stopped.connect(stop_cb)
    speech_recognizer.canceled.connect(stop_cb)

    # Start continuous speech recognition
    result=speech_recognizer.start_continuous_recognition()
  
    
    while not done:
        time.sleep(.5)

    
    return
Tags:

2

Answers


  1. 0
    • Here I was able to write the converted text to a file just by opening it using python inbuild function and writing the string data in it
    import  os
import  azure.cognitiveservices.speech  as  speechsdk

def  speechToText():
    speech_config = speechsdk.SpeechConfig('< KEY >', '< REGION>')
    speech_config.speech_recognition_language="en-US"

    audio_config = speechsdk.audio.AudioConfig(filename='PATH TO .WAV FILE')

    speech_recognizer = speechsdk.SpeechRecognizer(speech_config=speech_config, audio_config=audio_config)
    
    print("Starting the conversion")

    speech_recognition_result = speech_recognizer.recognize_once_async().get()
    
    // w+ is the read and write permission
    file1 = open("< PATH TO TEXT FILE >","w+")
    
    if speech_recognition_result.reason == speechsdk.ResultReason.RecognizedSpeech:
        file1.write(speech_recognition_result.text)
        file1.close()
        print("Recognized: {}".format(speech_recognition_result.text))

speechToText()

    Console Output :

    enter image description here

    Text File :

    enter image description here

    Login or Signup to reply.
  2. 0

    In the recognized_cb() function if condition "speechsdk.ResultReason.RecognizingSpeech == evt.result.reason" needs to be corrected to "speechsdk.ResultReason.RecognizedSpeech==evt.result.reason".
    Please make the following code changes so that you can see the desired results.

            if speechsdk.ResultReason.RecognizedSpeech==evt.result.reason and len(evt.result.text) > 0 :
            print('RECOGNIZED:', evt.result.text)
            output_file.write(evt.result.text)
            output_file.flush()
    Login or Signup to reply.
Please signup or login to give your own answer.

Back To Top
Search