Wit AI speech recognition python

Question

The code i wrote does not give the answer i want.

from wit import Wit

client = Wit("XXXYYYZZZ")

with open('sa.mp3', 'rb') as f:
  resp = client.speech(f, {'Content-Type': 'audio/wav'})
print('Yay, got Wit.ai response: ' + str(resp))`

the code gives me the "Yay, got Wit.ai response: {'entities': {}, 'intents': [], 'text': '', 'traits': {}}"

why?

`sa.mp3` is not a WAV file. You need to use the proper format identification. — Tim Roberts, Jan 19 '23 at 18:10
Try a PCM audio file since that is the default format that Wit expects. — iohans, Jan 19 '23 at 18:10

score 0 · Accepted Answer · edited Jan 19 '23 at 18:47

0

The most likely reason for your errored response is that you are sending the wrong format to Wit. Either change the format of the file to something like sa.wav or change the Content-Type to audio/mpeg.

PCM (audio/raw) Default format.
mp3 (audio/mpeg)
wav (audio/wav)

from wit import Wit

client = Wit("XXXYYYZZZ")

with open('sa.mp3', 'rb') as f:
  resp = client.speech(f, {'Content-Type': 'audio/mpeg'})
print('Yay, got Wit.ai response: ' + str(resp))

Docs: https://wit.ai/docs/http/20221114/

edited Jan 19 '23 at 18:47

Charles Duffy

280,126
43
390
441

answered Jan 19 '23 at 18:15

iohans

838
1
7
15

1

Gotta be careful putting four-space-indented code right below a bulleted list -- the indentation for the code can be interpreted as just being regular text inside the list element. – Charles Duffy Jan 19 '23 at 18:47

Wit AI speech recognition python

1 Answers1