Speech Recognition Low Accuracy Problem

[Uzair Lakhani]

Dear All Members,
I am using the below code for performing speech recognition using microphone:

Configuration configuration = new Configuration();

    // Set path to acoustic model.
    configuration.setAcousticModelPath("resource:/edu/cmu/sphinx/models/acoustic/WSJ_8gau_13dCep_16k_40mel_130Hz_6800Hz");

    // Set path to dictionary.
    configuration.setDictionaryPath("resource:/edu/cmu/sphinx/models/acoustic/WSJ_8gau_13dCep_16k_40mel_130Hz_6800Hz/dict/cmudict.0.6d");

    // Set language model.
    configuration.setLanguageModelPath("resource:/edu/cmu/sphinx/models/language/en-us.lm.dmp");

    LiveSpeechRecognizer recognizer = new LiveSpeechRecognizer(configuration);
    // Start recognition process pruning previously cached data.
    System.out.println("START");
    recognizer.startRecognition(true);
    SpeechResult result = recognizer.getResult();

    System.out.println("MESSAGE:");
    System.out.println(result.getHypothesis());

    // Pause recognition process. It can be resumed then with startRecognition(false).
    System.out.println("STOP");
    recognizer.stopRecognition();

    while(true){
        System.out.println("START");
        recognizer.startRecognition(true);
        result = recognizer.getResult();

        System.out.println("MESSAGE:");
        System.out.println(result.getHypothesis());

        // Pause recognition process. It can be resumed then with startRecognition(false).
        System.out.println("STOP");
        recognizer.stopRecognition();
    }

...
The problem is that I am not getting the text what I am saying in the microphone. Can you members guide me to increase the accuracy?

Thanks,
Uzair Lakhani

[Uzair Lakhani]

Dear All,

Here is some additional information.

My OS is Windows 7 Ultimate 64-bit. I assigned 2GB to the program because without assigning this memory the Out of Memory exception occurs. Additionally I am using the 5prealpha version of Sphinx.

I am giving input on Microphone and I am not a native English speaker. I want to input one to two words at once but not the whole sentence.

I am also getting difficulty in selecting the language model, dictionary, and acoustic model. Currently I am using the models mentioned in the last post.

Thanks,

[Alexander Solovets]

Your main loop is wrong, here is how it should be
https://github.com/cmusphinx/sphinx4/blob/58898035b02f0409d03cb2482030ad8d5e7ba0d3/sphinx4-samples/src/main/java/edu/cmu/sphinx/demo/transcriber/TranscriberDemo.java#L55

Sincerely, Alexander