Блог

Как я могу найти вероятность среднего вектора в средних векторах? Ниже приведены мои коды. Я использую пакет comirva.

 public class SpeechRecognition {
    private PointList list;
    public Vector<double[]> getMFCC(File file) {
        AudioInputStream audioInputStream;
        AudioPreProcessor preProcessor;
        MFCC ccm;
        Vector<double[]> features = new Vector<double[]>();
        try {
            audioInputStream = AudioSystem.getAudioInputStream(file);

            preProcessor = new AudioPreProcessor(audioInputStream, audioInputStream.getFormat().getSampleRate());
            ccm = new MFCC(preProcessor.getSampleRate());
            features = ccm.process(preProcessor);
            return features;
        } catch (IOException e) {

        } catch (UnsupportedAudioFileException e) {

        }
        return features;
    }

    public double[] getSingleFileFeatures(Vector<double[]> doubles) {
        double[] result = new double[600];
        int k = 0;
        for (int i = 0; i < doubles.size(); i  ) {
            for (int j = 0; j < doubles.get(0).length; j  ) {
                result[k] = doubles.get(i)[j];
                k  ;
            }
        }
        return resu<
    }

    public GaussianComponent getComponent(File file) {
        GaussianComponent component;
        double[] result = getSingleFileFeatures(getMFCC(file));
        PointList list = new PointList(result.length);
        list.add(result);
        KMeansClustering clustering = new KMeansClustering(list.size(), list);

        clustering.run();
        System.out.println(clustering.getClusterWeight(0));
        component = new GaussianComponent(clustering.getClusterWeight(0), clustering.getMean(0), clustering.getFullCovarianceMatrix(0));
        return component;
    }

    public GaussianMixture getGaussianMixtures(File[] files) {

        GaussianMixture mixture;
            getPointList(files);

            System.out.println();
            KMeansClustering clustering = new KMeansClustering(list.size(), list);

            clustering.run();
        for(int i=0; i<files.length; i  ){
            System.out.println(clustering.getClusterWeight(i));
        }
            mixture = new GaussianMixture(clustering.getClusterWeights(), clustering.getMeans(), clustering.getFullCovariances());


        return mixture;
    }
    public Matrix getMFCCAsMatrix(File file){
        return new Matrix(getMFCC(file).toArray(new double[0][]));
    }
    public PointList pointList(File file){
        double[] result = getSingleFileFeatures(getMFCC(file));
        PointList list = new PointList(600);
        list.add(result);

        return list;
    }
    public PointList getPointList(File[] files){
          list = new PointList(600);
          for(int i=0; i<files.length; i  ){
              double[] result = getSingleFileFeatures(getMFCC(files[i]));
              list.add(result);
          }
          System.out.println(list.size());
          return list;
    }

    public Matrix getPointListMatrix(File file){
        return pointList(file).get(0);
    }
}