Diarization - #1556
Conversation
Sorry, something went wrong.
|
@nnegrey , @dizcology , @theacodes Can you please review this PR? |
Sorry, something went wrong.
|
Looks like these use the speech_v1p1beta1 library, can we put these snippets into the beta_snippets.py file? |
Sorry, something went wrong.
| python transcribe_diarization.py \ | ||
| resources/Google_Gnome.wav | ||
| python transcribe_diarization.py \ | ||
| gs://cloud-ml-api-e2e-testing/speech/stereo_audio.wav |
There was a problem hiding this comment.
cloud-ml-api-e2e-testing public bucket?
Sorry, something went wrong.
|
Good catch @nnegrey. I moved all the 4 beta samples to beta_snippets.py and simplified a few things. |
Sorry, something went wrong.
| print('First alternative of result {}: {}' | ||
| .format(i, alternative.transcript)) | ||
| print('Speaker Tag for the first word: {}' | ||
| .format(alternative.words[0].speaker_tag)) |
There was a problem hiding this comment.
does each word in alternative.words have other relevant information concerning diarization besides speaker_tag? if so print them as well, if not please ignore this comment.
Sorry, something went wrong.
| os.path.join(RESOURCES, 'Google_Gnome.wav')) | ||
| out, err = capsys.readouterr() | ||
|
|
||
| assert 'OK Google stream stranger things from Netflix to my TV' in out |
There was a problem hiding this comment.
are there two different speakers in this audio file? if so assert something about the speaker_tags being returned correctly.
Sorry, something went wrong.
The samples in this code demo diarization: Basically who said what (when there is more than just one person talking).