Can we add acoustic model using SAPI RRS feed

  • Question

  • Hi

    I'm working on a project to develop a speech recognition application for dictation purpose for URDU language. I'm using system.speech.recognition namespace. As there is no current Urdu speech recognition engine so i'm using English speech recognition engine. And i'm able to recognize and display some words of Urdu language by adding pronunciation for each word of Urdu to be recognized in grammar file. But most of the time the wrong work has been recognized and i want to improve that. I want to know if there's any way that i can train existing acoustic model with my new language or i can add my own acoustic model. What should to improve that. I need suggestions.

    I'd appreciate if anyone could help me with this.   


    Sunday, December 7, 2014 1:48 PM