Azure Real time speech to text not working through python but curl is working

Question

We have enabled Azure speech to text service with private end point, when we try to use below curl command it we able to get output

curl -i --location 'https://xxxxxxxxxxx?language=en-US' --header 'Accept: application/json' --header 'Ocp-Apim-Subscription-Key: xxxxxxxxx' --header 'Content-Type: audio/wav' --data-binary '@/app/temp/longer_audio.wav'

but when used attached script it is event is getting cancelled, can anyone please help here

Answer 1

Hi Ulhas Hulyal, Nilesh

we are able reproduce the issue and able to solve in our environment can you please check once with

import time
import azure.cognitiveservices.speech as speechsdk
 
def conversation_transcriber_recognition_canceled_cb(evt: speechsdk.SessionEventArgs):
    print('Canceled event')
 
def conversation_transcriber_session_stopped_cb(evt: speechsdk.SessionEventArgs):
    print('SessionStopped event')
 
def conversation_transcriber_transcribed_cb(evt: speechsdk.SpeechRecognitionEventArgs):
    print('\nTRANSCRIBED:')
    if evt.result.reason == speechsdk.ResultReason.RecognizedSpeech:
        print('\tText={}'.format(evt.result.text))
        print('\tSpeaker ID={}\n'.format(evt.result.speaker_id))
    elif evt.result.reason == speechsdk.ResultReason.NoMatch:
        print('\tNOMATCH: Speech could not be TRANSCRIBED: {}'.format(evt.result.no_match_details))
 
def conversation_transcriber_transcribing_cb(evt: speechsdk.SpeechRecognitionEventArgs):
    print('TRANSCRIBING:')
    print('\tText={}'.format(evt.result.text))
    print('\tSpeaker ID={}'.format(evt.result.speaker_id))
 
def conversation_transcriber_session_started_cb(evt: speechsdk.SessionEventArgs):
    print('SessionStarted event')
 
def recognize_from_file():
    # This example requires environment variables named "SPEECH_KEY" and "SPEECH_REGION"
    speech_config = speechsdk.SpeechConfig(subscription="<KEY>", region="<REGION>")
    speech_config.speech_recognition_language="en-US"
    speech_config.set_property(property_id=speechsdk.PropertyId.SpeechServiceResponse_DiarizeIntermediateResults, value='true')
 
    audio_config = speechsdk.audio.AudioConfig(filename=r"FILE_PATH")
    conversation_transcriber = speechsdk.transcription.ConversationTranscriber(speech_config=speech_config, audio_config=audio_config)
 
    transcribing_stop = False
 
    def stop_cb(evt: speechsdk.SessionEventArgs):
        #"""callback that signals to stop continuous recognition upon receiving an event `evt`"""
        print('CLOSING on {}'.format(evt))
        nonlocal transcribing_stop
        transcribing_stop = True
 
    # Connect callbacks to the events fired by the conversation transcriber
    conversation_transcriber.transcribed.connect(conversation_transcriber_transcribed_cb)
    conversation_transcriber.transcribing.connect(conversation_transcriber_transcribing_cb)
    conversation_transcriber.session_started.connect(conversation_transcriber_session_started_cb)
    conversation_transcriber.session_stopped.connect(conversation_transcriber_session_stopped_cb)
    conversation_transcriber.canceled.connect(conversation_transcriber_recognition_canceled_cb)
    # stop transcribing on either session stopped or canceled events
    conversation_transcriber.session_stopped.connect(stop_cb)
    conversation_transcriber.canceled.connect(stop_cb)
 
    conversation_transcriber.start_transcribing_async()
 
    # Waits for completion.
    while not transcribing_stop:
        time.sleep(.5)
 
    conversation_transcriber.stop_transcribing_async()
 
# Main
 
try:
    recognize_from_file()
except Exception as err:
    print("Encountered exception. {}".format(err))

output.

I hope this helps you. Thank you.