Closed joelprince25 closed 5 years ago
I'm not sure I've used that feature before. Do you have a sample input/output file?
Here's a sample input and output file. In this case, the service didn't successfully identify two channels but the structure of the output json wouldn't change. channel_example.zip
Thanks, Joel. It looks like the same text is happening on both channels? Not sure I understand how this works.
I can't share the audio files I'm working with because of confidentiality reasons. This sample file has the same voice on both channels so that's why it's the same. In reality, different channels will have different sounds, say, an agent and a customer will have different channels. I'm having some trouble finding good sample files to send you.
I don't really need the sample audio files. Can you provide me with a sample output? Can a channel have speaker identification, too?
Thanks!
Just created a pull request. No a channel cannot have speaker identification. AWS Transcribe let's you do either channel identification or speaker identification.
On Mon, Feb 11, 2019 at 4:54 PM Jason Purdy notifications@github.com wrote:
I don't really need the sample audio files. Can you provide me with a sample output? Can a channel have speaker identification, too?
Thanks!
— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub https://github.com/purdy/aws-transcribe-transcript/issues/4#issuecomment-462561955, or mute the thread https://github.com/notifications/unsubscribe-auth/AHw-f7pfdDBcvxYE6pW8JiWelSEmxyK3ks5vMhCqgaJpZM4Z_36N .
Merged in ... thanks, Joel!
Amazon Transcribe now enables channel identification. I was wondering if you were working on a similar script to make the channel output identification script more readable.