Speech recognition using .NET Frameworks as a web service

Question

If I want to build a web application to be used as a service for speech recognition. Simply, I want to transfer a speech to my website to be recognized and then get the speech as a text.

What is the best way to implement this application to run on a ASP.NET website?

Answer 1

One way to do this is run a WCF service on your server to do the Speech recognition using Microsoft Speech Recognition API (SAPI) and let your web service to talk with your WCF Service . Basically, the web service needs to get the audio, stream it and send it to the WCF service.

YOu can find more information on this post: http://blog.renzuocheng.com/2011/03/speech-recognitionsapi-in-asp-net2/

Answer 2

There are many parts here.

• How to read the user and send it to the server
• How to make speech recognition
• Return the result to the user

Here I answer to the first part, how to read the user input. For how to make speech recognition, I think that you need to make a research for find the correct module. One that I know is the Microsoft .NET Speech SDK 1 that you need to find it.

Using HTML5

A new way is to use the x-webkit-speech key as:

<input type="text" x-webkit-speech="x-webkit-speech" />

You then can send this data to the server for speech recognition.

Some examples:

• Stack Overflow question x-webkit-speech help

• How to Use HTML5 Speech Input Fields

• x-webkit-speech input and textareas

Using Flash

To be able to use a speech recognition, you need to add Flash to your site to be able to read the microphone from the user computer (this is done with the permission of the user).

On your script on the Flash, you get what the user speaks and your processing on the server side.