Speech Recognition for Own Language RRS feed

  • Question

  • I want to develop my own speech recognition software in for punjabi Language (India-Punjabi)

    for reference

    I read about Microsoft speech recognition libraries but my requirement is to use my own training set, acoustic model and grammar. 

    I already implemented sphinx speech recognition ( but it is in java.

    How it is possible?

    Microsoft Customer Support Forms

    Friday, December 29, 2017 12:07 PM


All replies

  • Hello Aman,

    Thank you for your post.

    Based on your description, I am afraid that the issue is out of support of C# General Question forum which mainly discuss and ask the C# programming language, IDE, libraries and tools. I'm not familiar with speech recognition algorithms or implementation mechanism. This belong to the category of artificial intelligence.

    I suggest that you can consult your issue directly on speech recognition forum for better solution and support.

    If you have any issues when you use C#, please feel free to let me know.

    Best regards,

    Neil Hu

    MSDN Community Support
    Please remember to click "Mark as Answer" the responses that resolved your issue, and to click "Unmark as Answer" if not. This can be beneficial to other community members reading this thread. If you have any compliments or complaints to MSDN Support, feel free to contact

    Tuesday, January 2, 2018 9:38 AM
  • Hi Aman,

    It looks like a homework to me. We won't do your homework but we can show you the path in which direction to approach.

    Have you look into speech recognizer class in .NET. It has a instruction how to use this class in C#, you can pass your dataset, grammar here. Please look into below articles for more information on how to use this class. Hope this helps you.

    Get Started with Speech Recognition

    Speech recognition, speech to text, text to speech, and speech synthesis in C#

    Voice Recognition : Speech Recognition with .NET Desktop Applications

    Sabah Shariq

    [If a post helps to resolve your issue, please click the "Mark as Answer" of that post or click Answered "Vote as helpful" button of that post. By marking a post as Answered or Helpful, you help others find the answer faster. ]

    Tuesday, January 2, 2018 11:51 AM
  • sir, it is not a homework to you, i am just asking to experts... is there any possibility to train own acoustic model for speech recognition.. 

    links you sent me that work only for already trained model. i can't train model for own cultural language.


    Microsoft Customer Support Forms

    Tuesday, September 11, 2018 11:02 AM