Multi Channel #1561

happyhuman · 2018-07-12T21:52:09Z

Implemented the multi channel sample

happyhuman · 2018-07-12T21:57:06Z

@nnegrey , @dizcology , @tswast can you please review this PR?

tswast · 2018-07-13T16:42:45Z

speech/cloud-client/beta_snippets.py

@@ -157,6 +158,35 @@ def transcribe_file_with_diarization(path):
 # [END speech_transcribe_diarization]


+# [START speech_transcribe_multichannel]


Nit: per our snippet rubric snippets should show concrete values for variables whenever possible. I recommend moving this region tag into the function rather than outside of it. The snippet should also show the necessary imports.

Example:

https://github.com/GoogleCloudPlatform/google-cloud-python/blob/b022ca9d2176411c1d0a528da9ba820d24f0630e/docs/bigquery/snippets.py#L180-L181

Note: we comment out the import and client creation in BigQuery because we have ~100 snippets and repeating auth every time slows down the tests substantially, but this is not necessary if there are only a few snippets as there are here.

Good observation @tswast. However all other samples in the speech directory seem to have the region tags outside the functions. So, if we move the region tags inside the function for this sample, it will look inconsistent with the rest of the samples for speech. What do you think?

Better to follow the new standard than to follow the old one. Eventually we will convert all samples to the new standard (starting with translate).

Fair point @tswast, I moved the region tags in beta_snippets.py inside the functions

tswast · 2018-07-13T16:46:21Z

speech/cloud-client/beta_snippets.py

@@ -157,6 +158,35 @@ def transcribe_file_with_diarization(path):
 # [END speech_transcribe_diarization]


+# [START speech_transcribe_multichannel]
+def transcribe_file_with_multichannel(speech_file):


I recommend showing an example value for speech_file within the snippet.

# [START speech_transcribe_multichannel] from google.cloud import speech client = speech.SpeechClient() # TODO(developer): Uncomment and set to a path to your audio file. # speech_file = 'path/to/file.wav' ...

Example:

https://github.com/GoogleCloudPlatform/google-cloud-python/blob/b022ca9d2176411c1d0a528da9ba820d24f0630e/docs/bigquery/snippets.py#L183

Good call. I added it (and also for Diarization sample which was merged in yesterday).

tswast

A couple nits. LGTM after they are addressed (either by renaming the argument or updating the comment)

tswast · 2018-07-13T22:43:08Z

speech/cloud-client/beta_snippets.py

    client = speech.SpeechClient()

+    # TODO(developer): Uncomment and set to a path to your audio file.
+    # speech_file = 'path/to/file.wav'


In this sample the variable appears to be called "path".

tswast · 2018-07-13T22:43:16Z

speech/cloud-client/beta_snippets.py

    client = speech.SpeechClient()

+    # TODO(developer): Uncomment and set to a path to your audio file.
+    # speech_file = 'path/to/file.wav'


In this sample the variable appears to be called "path".

tswast · 2018-07-13T22:43:22Z

speech/cloud-client/beta_snippets.py

    client = speech.SpeechClient()

+    # TODO(developer): Uncomment and set to a path to your audio file.
+    # speech_file = 'path/to/file.wav'


In this sample the variable appears to be called "path".

tswast · 2018-07-13T22:43:26Z

speech/cloud-client/beta_snippets.py

    client = speech.SpeechClient()

+    # TODO(developer): Uncomment and set to a path to your audio file.
+    # speech_file = 'path/to/file.wav'


In this sample the variable appears to be called "path".

dizcology · 2018-07-13T23:23:28Z

speech/cloud-client/beta_snippets.py

    """Transcribe the given audio file using an enhanced model."""
+    # [START speech_transcribe_file_with_enhanced_model]


I see this is a different way to use the region tags than commonly done in Python so far. Is this the convention going forward?

Yes, this is a result of our most recent samples rubric working group. I'll be presenting on this in a team meeting post-Next.

dizcology · 2018-07-13T23:26:28Z

speech/cloud-client/beta_snippets.py

+        print('-' * 20)
+        print('First alternative of result {}'.format(i))
+        print(u'Transcript: {}'.format(alternative.transcript))
+        print(u'Channel Tag: {}'.format(result.channel_tag))


what happens where there are multiple channels? do we get the output from all channels into the same alternative? or does the output from the second channel go to the second alternative, and so on?

dizcology · 2018-07-13T23:26:55Z

speech/cloud-client/beta_snippets_test.py

+        os.path.join(RESOURCES, 'Google_Gnome.wav'))
+    out, err = capsys.readouterr()
+
+    assert 'OK Google stream stranger things from Netflix to my TV' in out


this is a single channel audio file, use an actual multichannel test file.

* Implemented the multi channel sample * Added parameter comment * Moved region tags inside the functions * Deleted the extra line * Fixing typos

Implemented the multi channel sample

bbb8a5c

googlebot added the cla: yes This human has signed the Contributor License Agreement. label Jul 12, 2018

happyhuman requested review from dizcology, tswast and nnegrey July 12, 2018 21:56

tswast reviewed Jul 13, 2018

View reviewed changes

happyhuman added 3 commits July 13, 2018 14:05

Added parameter comment

a8021d4

Moved region tags inside the functions

bc7272f

Deleted the extra line

02d13ed

tswast approved these changes Jul 13, 2018

View reviewed changes

Fixing typos

b165d37

tswast approved these changes Jul 13, 2018

View reviewed changes

happyhuman merged commit 1fe1ca8 into master Jul 13, 2018

happyhuman deleted the multichannel branch July 13, 2018 23:17

dizcology reviewed Jul 13, 2018

View reviewed changes

telpirion pushed a commit that referenced this pull request Jan 13, 2023

Multi Channel [(#1561)](#1561)

80f9dc6

* Implemented the multi channel sample * Added parameter comment * Moved region tags inside the functions * Deleted the extra line * Fixing typos

telpirion mentioned this pull request Jan 13, 2023

chore(speech): migrate code from googleapis/python-speech #8982

Merged

8 tasks

dandhlee pushed a commit that referenced this pull request Feb 9, 2023

Multi Channel [(#1561)](#1561)

074b456

* Implemented the multi channel sample * Added parameter comment * Moved region tags inside the functions * Deleted the extra line * Fixing typos

telpirion pushed a commit that referenced this pull request Mar 13, 2023

Multi Channel [(#1561)](#1561)

991d55e

* Implemented the multi channel sample * Added parameter comment * Moved region tags inside the functions * Deleted the extra line * Fixing typos

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Multi Channel #1561

Multi Channel #1561

happyhuman commented Jul 12, 2018

happyhuman commented Jul 12, 2018

tswast Jul 13, 2018

happyhuman Jul 13, 2018

tswast Jul 13, 2018

happyhuman Jul 13, 2018

tswast Jul 13, 2018

happyhuman Jul 13, 2018

tswast left a comment

tswast Jul 13, 2018

tswast Jul 13, 2018

tswast Jul 13, 2018

tswast Jul 13, 2018

dizcology Jul 13, 2018

tswast Jul 13, 2018

dizcology Jul 13, 2018

dizcology Jul 13, 2018

		@@ -157,6 +158,35 @@ def transcribe_file_with_diarization(path):
		# [END speech_transcribe_diarization]


		# [START speech_transcribe_multichannel]

		"""Transcribe the given audio file using an enhanced model."""
		# [START speech_transcribe_file_with_enhanced_model]

Multi Channel #1561

Multi Channel #1561

Conversation

happyhuman commented Jul 12, 2018

happyhuman commented Jul 12, 2018

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

tswast left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment