Results 1 to 4 of 4

Thread: Provide a URL for SpeechRecognizer audio parameter

  1. #1
    Junior Member
    Join Date
    Feb 2012
    Posts
    6

    Question Provide a URL for SpeechRecognizer audio parameter

    Is it possible to provide a URL for the audio parameter instead of streaming a file to the server?

  2. #2
    Administrator
    Join Date
    Mar 2011
    Posts
    230
    We've tried to simplify the SDK as much as possible by handling the entire network transaction for the developer.

    If you require more flexibility in your application, you should consider looking at our direct API: http://www.ispeech.org/api

    The API gives you direct access to the audio (in multiple formats) and allows for server implementations.

  3. #3
    Junior Member
    Join Date
    Feb 2012
    Posts
    6
    Thanks for your reply, Florencio.

    I don't think I defined my use case very accurately. I am using the iSpeech API, but I can't use it in the way you suggest. Here is the use case:

    Speech-To-Text use case for my voice app, which I will call MyApp.


    - MyApp app uses a 3rd party telephony API (Twilio) to record short segments of speech.
    - Twilio stores the recorded audio on its servers and returns a URI to MyApp for the audio file.
    - * MyApp downloads the audio file at the URI identified by Twilio to MyApp server
    - * Using the iSpeech API, MyApp streams the audio file up to iSpeech, along with a command dictionary.
    - * iSpeech responds with converted text
    - MyApp takes converted text, does stuff with it
    - Process starts over at the top

    I know that iSpeech could be doing the whole thing round-trip, but that's not my use case. I have to use Twilio and may need to use other services as well, Tropo for example. I put stars (*) next to those steps of the use case that would benefit from the feature request.

    Right now, MyApp downloads an audio file, only to turn around and stream it right back to iSpeech. It is desirable to be able to *optionally* provide iSpeech with the URI for an audio file, which the iSpeech server could retrieve and process. For iSpeech the load would be the same - it still needs to obtain and process a single stream of audio.

    Thanks for your further consideration.

    Ed

  4. #4
    Administrator
    Join Date
    Mar 2011
    Posts
    230
    I see. I will add that as one of the feature requests, you might see that in the up and coming future.

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •