ShazamSample.java

package org.gennai.audio_analyzer.analysis;

import java.io.IOException;

import javax.sound.sampled.LineUnavailableException;
import javax.sound.sampled.UnsupportedAudioFileException;

import org.gennai.audio_analyzer.analysis.AudioTest.Audio;
import org.gennai.audio_analyzer.analysis.audio.AudioSamples;
import org.gennai.audio_analyzer.analysis.config.FingerprintConfig;
import org.gennai.audio_analyzer.analysis.dao.IModelReference;
import org.gennai.audio_analyzer.analysis.dao.data.SoundData;
import org.gennai.audio_analyzer.analysis.inmemory.InMemoryModelService;
import org.gennai.audio_analyzer.analysis.inmemory.RAMStorage;

public class ShazamSample {

  private FingerprintConfig fingerprintConfig;
  private InMemoryModelService modelService;
  private Analyzer analyzer;

  private ShazamSample() {
    // 設定
    fingerprintConfig = new FingerprintConfig();
    // 4に固定
    fingerprintConfig.getSpectrogramConfig().setLogBins(4);
    /*
    fingerprintConfig.setNormalizeSignal(false);
    fingerprintConfig.getSpectrogramConfig().setWdftSize(2048);
    fingerprintConfig.getSpectrogramConfig().setOverlap(64);
    */

    RAMStorage storage = new RAMStorage(
        fingerprintConfig.getHashingConfig().getNumberOfLSHTables());
    modelService = new InMemoryModelService(storage);

    analyzer = new Analyzer();
  }

  private void insertHashData(String wavFile)
      throws UnsupportedAudioFileException, IOException, LineUnavailableException {
    // WAVEファイル読み込み
    Audio audio = AudioTest.readWave("../samples4/nom1.wav");

    AudioSamples audioSamples = new AudioSamples(audio.getData(), audio.getOrigin(),
        (int) audio.getFormat().getSampleRate());

    // ハッシュを作成
    long[] hashes = analyzer.spectrogram(audioSamples, fingerprintConfig.getSpectrogramConfig());

    // メタデータをモデルに登録
    double length = (double) audio.getData().length / audio.getFormat().getSampleRate()
        / audio.getFormat().getChannels();
    IModelReference<?> soundReference = modelService.saveSound(new SoundData(length));

    // ハッシュをモデルに登録
    modelService.insertDataPointsForSound(hashes, soundReference);
  }

  private void query(String wavFile)
      throws UnsupportedAudioFileException, IOException, LineUnavailableException {
    // WAVEファイル読み込み
    Audio audio = AudioTest.readWave("../samples4/nom1.wav");

    AudioSamples audioSamples = new AudioSamples(audio.getData(), audio.getOrigin(),
        (int) audio.getFormat().getSampleRate());

    // ハッシュを作成
    long[] hashes = analyzer.spectrogram(audioSamples, fingerprintConfig.getSpectrogramConfig());

    // モデルを検索
    MatchResult ret = analyzer.query(hashes, modelService);
    // ヒットしたか？
    if (ret.containsMatches()) {
      // ベストマッチ
      System.out.printf("best id: %d, score: %d\n", ret.getBestMatch().getSoundRefrence().getId(),
          ret.getBestMatch().getMostOffsetCount());

      // マッチリスト
      for (MatchEntry entry : ret.getMatchEntries()) {
        System.out.printf("id: %d, score: %d\n", entry.getSoundRefrence().getId(),
            entry.getMostOffsetCount());
      }
    } else {
      System.out.println("no match");
    }
  }

  public static void main(String[] args)
      throws UnsupportedAudioFileException, IOException, LineUnavailableException {
    ShazamSample shazamSample = new ShazamSample();
    shazamSample.insertHashData("../samples4/nom1.wav");
    shazamSample.query("../samples4/nom1.wav");
  }
}