You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
PaddleSpeech/runtime/examples/vad/vad-android-demo/README.md

4.7 KiB

This VAD library can process audio in real-time utilizing Gaussian Mixture Model (GMM) which helps identify presence of human speech in an audio sample that contains a mixture of speech and noise. VAD work offline and all processing done on device.

Library based on WebRTC VAD from Google which is reportedly one of the best available: it's fast, modern and free. This algorithm has found wide adoption and has recently become one of the gold-standards for delay-sensitive scenarios like web-based interaction.

If you are looking for a higher accuracy and faster processing time I recommend to use Deep Neural Networks(DNN). Please see for reference the following paper with DNN vs GMM comparison.

drawing

Parameters

VAD library only accepts 16-bit mono PCM audio stream and can work with next Sample Rates, Frame Sizes and Classifiers.

 
Valid Sample Rate Valid Frame Size
8000Hz 80, 160, 240
16000Hz 160, 320, 480
32000Hz 320, 640, 960
48000Hz 480, 960, 1440
 
Valid Classifiers
NORMAL
LOW_BITRATE
AGGRESSIVE
VERY_AGGRESSIVE

Silence duration (ms) - this parameter used in Continuous Speech detector, the value of this parameter will define the necessary and sufficient duration of negative results to recognize it as silence.

Voice duration (ms) - this parameter used in Continuous Speech detector, the value of this parameter will define the necessary and sufficient duration of positive results to recognize result as speech.

Recommended parameters:

  • Sample Rate - 16KHz,
  • Frame Size - 160,
  • Mode - VERY_AGGRESSIVE,
  • Silence Duration - 500ms,
  • Voice Duration - 500ms;

Usage

VAD supports 2 different ways of detecting speech:

  1. Continuous Speech listener was designed to detect long utterances without returning false positive results when user makes pauses between sentences.
 Vad vad = new Vad(VadConfig.newBuilder()
                .setSampleRate(VadConfig.SampleRate.SAMPLE_RATE_16K)
                .setFrameSize(VadConfig.FrameSize.FRAME_SIZE_160)
                .setMode(VadConfig.Mode.VERY_AGGRESSIVE)
                .setSilenceDurationMillis(500)
                .setVoiceDurationMillis(500)
                .build());

        vad.start();
        
        vad.addContinuousSpeechListener(short[] audioFrame, new VadListener() {
            @Override
            public void onSpeechDetected() {
                //speech detected!
            }

            @Override
            public void onNoiseDetected() {
                //noise detected!
            }
        });
        
        vad.stop();
  1. Speech detector was designed to detect speech/noise in small audio frames and return result for every frame. This method will not work for long utterances.
 Vad vad = new Vad(VadConfig.newBuilder()
                .setSampleRate(VadConfig.SampleRate.SAMPLE_RATE_16K)
                .setFrameSize(VadConfig.FrameSize.FRAME_SIZE_160)
                .setMode(VadConfig.Mode.VERY_AGGRESSIVE)
                .build());

        vad.start();
        
        boolean isSpeech = vad.isSpeech(short[] audioFrame);
        
        vad.stop();

Requirements

Android VAD supports Android 4.1 (Jelly Bean) and later.

Development

To open the project in Android Studio:

  1. Go to File menu or the Welcome Screen
  2. Click on Open...
  3. Navigate to VAD's root directory.
  4. Select setting.gradle

Download

Gradle is the only supported build configuration, so just add the dependency to your project build.gradle file:

  1. Add it in your root build.gradle at the end of repositories:
allprojects {
   repositories {
     maven { url 'https://jitpack.io' }
   }
}
  1. Add the dependency
dependencies {
    implementation 'com.github.gkonovalov:android-vad:1.0.1'
}

You also can download precompiled AAR library and APK files from GitHub's releases page.


Georgiy Konovalov 2021 (c) MIT License