SPEECH_ACTIVITY_END event not returned for Japanese audio in long model. [370627603]

Assigned

Feature Request

Status Update

No update yet.

Description

si...@google.com

created issue #1

Oct 1, 2024 03:30AM

Description: Voice activity timeout setting ‘Speech end timeout’ [1] is getting violated since Speech_activity_end is not returned for Japanese Audio while using the ‘long’ model on streaming mode. However, this seems to be working fine for EN audio.

Intended behavior: SPEECH_ACTIVITY_END returned for JP audio as well in the long model.

Workaround: Using Telephony model instead seems to resolve the issue.

[1]

https://cloud.google.com/speech-to-text/v2/docs/reference/rpc/google.cloud.speech.v2#voiceactivitytimeout

Comments

va...@google.com <va...@google.com> Oct 1, 2024 06:37AM

Assigned to ar...@google.com.

ar...@google.com <ar...@google.com> #2Oct 1, 2024 05:14PM

Reassigned to gc...@google.com.

I have informed our engineering team of this feature request. There is currently no ETA for its implementation.

A current workaround would be to check the returned "boundingPoly" [1] "vertices" for the returned "textAnnotations". If the calculated rectangle's heights > widths, than your image is sideways.

[1]

https://cloud.google.com/vision/reference/rest/v1/images/annotate#boundingpoly

Issue 370627603

Description

Issue summary

Comments

va...@google.com <va...@google.com> Oct 1, 2024 06:37AM

ar...@google.com <ar...@google.com> #2Oct 1, 2024 05:14PM

Add comment

Issue metadata