2

When I run my code, wit.ai try to recognize English words of "myspeech.wav" but that is Persian voice, How should I change my code?

def RecognizeSpeech(AUDIO_FILENAME):
    audio = open('myspeech.wav','r')
    headers = {'authorization': 'Bearer ' + wit_access_token,
               'Content-Type': 'audio/wav'}
    resp = requests.post(API_ENDPOINT, headers = headers,
                     data = audio)
    data = json.loads(resp.content)
    text = data['_text']
    return text
Monako
  • 51
  • 4
  • For those downvoting: Wit.ai [supports Persian](https://wit.ai/faq), so this is just a question of "how to turn it on", not "how do I train an AI to recognize Persian?", as I initially thought it was. – Alex von Brandenfels Feb 26 '18 at 21:37

1 Answers1

5

If you set the language in the setting of your app in wit console to Persian, that should works!

You can alternatively use:

curl -XPOST 'https://api.wit.ai/speech' -i -L -H 'Authorization: Bearer YOURAPPTOKEN' -H "Content-Type: audio/wav" --data-binary "@sample.wav"

which returns the JSON of the recognized text, intent, and entities.

Ali Zarezade
  • 871
  • 9
  • 22