python - best approach towards cloud speech "streaming" solution? -


i making small transcription service , thinking best options, advice welcome.

i'm modifying google voice aiy demo long form (~1hr) transcripts. need store in google cloud storage bucket.

i have found transcription takes same time length of recording. (if incorrect, please tell me!). hour long recordings waiting around transcript finish long. figured start uploading google cloud storage in background.

what best strategy here?

  • should upload in chunks, , transcribe in background? mean chunks have overlap, won't cause duplications in transcript?

  • should "stream" directly google cloud storage? , not store locally?

  • should transcribe directly? how hour long sessions? (synchronous transcription google cloud speech under 1 min)

i want minimize waiting times transcript, , errors caused this.


Comments

Popular posts from this blog

angular - Ionic slides - dynamically add slides before and after -

minify - Minimizing css files -

Add a dynamic header in angular 2 http provider -