Speech recoginiton accuracy C#

Question

0.00/5 (No votes)

See more:

Hi,

I'm trying to recognize what user is speaking using microphone and compare it with the predefined text and validate. I have used the System.Speech dll.
I have doubts regarding the same.

1.Currently I am using grammar builder with hard-coded values in it and matching it with what user is saying.My Code below.
Problem with below code is that sometimes it matches words which sound similar
for example "hear" and "year". Also if I have word say "developer" and if user says "development" then it will pick "developer" from grammar. So is there way to avoid it or increase the accuracy of it?

2. Is there a way where I can record everything what user has said and then compare it? If yes, then how?

What I have tried:

C#

static void Main(string[] args)
      {
          SpeechRecognizer recognizer = new SpeechRecognizer();

          Choices colors = new Choices();
          colors.Add(new string[] { "red", "green", "blue" });

          GrammarBuilder gb = new GrammarBuilder();
          gb.Append(colors);

          Grammar g = new Grammar(gb);
          gb.Culture = Thread.CurrentThread.CurrentCulture;
          recognizer.LoadGrammar(g);

          recognizer.SpeechRecognized +=
            new EventHandler<SpeechRecognizedEventArgs>(sre_SpeechRecognized);
      }
      void sre_SpeechRecognized(object sender, SpeechRecognizedEventArgs e)
      {
          if (e.Result.Text == "red")
          {
              Console.WriteLine("Recognized red color");
          }
      }

Posted 16-Jun-16 2:47am

Member 12586674

Updated 16-Jun-16 4:12am

Add a Solution

1 solution

Add a Solution

Add your solution here

Treat my content as plain text, not as HTML

Preview 0

…

Existing Members

Sign in to your account

...or Join us

Download, Vote, Comment, Publish.

Your Email
Password
Forgot your password?

Your Email
This email is in use. Do you need your password?
Optional Password

I have read and agree to the Terms of Service and Privacy Policy
Please subscribe me to the CodeProject newsletters

When answering a question please:

Read the question carefully.
Understand that English isn't everyone's first language so be lenient of bad spelling and grammar.
If a question is poorly phrased then either ask for clarification, ignore it, or edit the question and fix the problem. Insults are not welcome.
Don't tell someone to read the manual. Chances are they have and don't get it. Provide an answer or move on to the next question.

Let's work to help developers, not make them feel stupid.

This content, along with any associated source code and files, is licensed under The Code Project Open License (CPOL)

Sergey Alexandrovich Kryukov · Accepted Answer · 2016-06-16T04:12:00

It has nothing to do with your code. This quality (fidelity) problem is a spectacular fallacy of the engine you are using. It does work if you pronounce everything very clearly and if your grammar is very small. Grow your grammar size a bit, introduce a couple of words which sounds even remotely similar, and… forget it. Ridiculously, the API includes dictation grammar, which apparently covers big part of the language thesaurus. This option is totally useless, no matter how good the user's pronunciation is.

It's noteworthy that, say, on Android you can install speech recognition as yet another keyboard. You can practically use speech input in dictation mode in several languages, with the fidelity which can really compete with on-screen virtual keyboards.

—SA