Hello I am trying to use the following sdk:
https://github.com/watson-developer-cloud/python-sdk/blob/master/examples/speech_to_text_v1.py
In order to get the text transcription of a wav file as follows:
import json
from os.path import join, dirname
from watson_developer_cloud import SpeechToTextV1
speech_to_text = SpeechToTextV1(
username='XXXXXXXXX',
password='XXXXXXXXX',
x_watson_learning_opt_out=False
)
print(json.dumps(speech_to_text.models(), indent=2))
print('I am using the spanish model for this test')
print(json.dumps(speech_to_text.get_model('es-ES_NarrowbandModel'), indent=2))
with open(join(dirname(__file__), '/Users/Downloads/python-sdk-master/examples/test.wav'),
'rb') as audio_file:
print(json.dumps(speech_to_text.recognize(
audio_file, content_type='audio/wav', timestamps=True,
word_confidence=True),
indent=2))
The problem is that every time that I run the request as follows:
python speech.py
I am getting the same result and it does not matter if I change the name of the file in the parameters I always get:
I am using the spanish model for this test
{
"name": "es-ES_NarrowbandModel",
"language": "es-ES",
"sessions": "https://stream.watsonplatform.net/speech-to-text/api/v1/sessions?model=es-ES_NarrowbandModel",
"url": "https://stream.watsonplatform.net/speech-to-text/api/v1/models/es-ES_NarrowbandModel",
"rate": 8000,
"supported_features": {
"custom_language_model": false,
"speaker_labels": true
},
"description": "Spanish narrowband model."
}
{
"results": [
{
"alternatives": [
{
"word_confidence": [
[
"yeah",
0.361
],
[
"and",
0.867
],
[
"on",
0.448
],
[
"the",
0.243
],
[
"loss",
0.172
],
[
"of",
0.68
],
[
"my",
0.953
],
[
"honor",
0.131
],
[
"and",
0.12
],
[
"sometimes",
0.23
],
[
"platter",
0.659
],
[
"and",
0.339
],
[
"also",
0.337
],
[
"got",
0.227
],
[
"asking",
0.383
],
[
"about",
0.1
],
[
"someone",
0.571
],
[
"economies",
0.144
],
[
"on",
0.146
],
[
"both",
0.093
]
],
"confidence": 0.368,
"transcript": "yeah and on the loss of my honor and sometimes platter and also got asking about someone economies on both ",
"timestamps": [
[
"yeah",
0.18,
0.47
],
[
"and",
0.72,
1.28
],
[
"on",
1.28,
1.41
],
[
"the",
1.41,
1.48
],
[
"loss",
1.48,
1.78
],
[
"of",
1.78,
1.89
],
[
"my",
1.89,
2.04
],
[
"honor",
2.04,
2.37
],
[
"and",
2.37,
2.53
],
[
"sometimes",
2.56,
3.17
],
[
"platter",
3.17,
3.53
],
[
"and",
4.04,
4.17
],
[
"also",
4.17,
4.45
],
[
"got",
4.45,
4.63
],
[
"asking",
4.63,
4.97
],
[
"about",
4.97,
5.18
],
[
"someone",
5.18,
5.45
],
[
"economies",
5.45,
5.97
],
[
"on",
5.97,
6.12
],
[
"both",
6.12,
6.34
]
]
}
],
"final": true
},
{
"alternatives": [
{
"word_confidence": [
[
"even",
0.547
],
[
"in",
0.586
],
[
"the",
0.766
],
[
"planet",
0.276
],
[
"of",
0.131
],
[
"my",
0.188
],
[
So I would like to appreciate support to overcome this task,
You shouldn't need the join(dirname(__file__), if you're specifying a fully qualified file path. If the path of your desired audio file isn't in a relative path, I'd try removing the join, although if that's the problem I'd expect a file not found error.