webrtcvad

README

A quick n' dirty Go port of py-webrtcvad Voice Activity Detector (VAD).

A VAD classifies a piece of audio data as being voiced or unvoiced. It can be useful for telephony and speech recognition.

The VAD that Google developed for the WebRTC project is reportedly one of the best available, being fast, modern and free.

Usage

Go-get the package. You don't need to have webrtc installed.

go get github.com/maxhawkins/go-webrtcvad

Feed raw audio samples into the VAD:

reader, err := wav.NewReader("test.wav")
if err != nil {
    log.Fatal(err)
}

vad, err := webrtcvad.New()
if err != nil {
    log.Fatal(err)
}

if err := vad.SetMode(2); err != nil {
    log.Fatal(err)
}

rate := 32000 // kHz
frame := make([]byte, 320*2)

if ok := vad.ValidRateAndFrameLength(rate, len(frame)); !ok {
    log.Fatal("invalid rate or frame length")
}
for {
    _, err := io.ReadFull(reader, frame)
    if err == io.EOF || err == io.ErrUnexpectedEOF {
        break
    }
    if err != nil {
        log.Fatal(err)
    }

    active, err := vad.Process(rate, frame)
    if err != nil {
        log.Fatal(err)
    }

    fmt.Println(active)
}

Documentation

Index

• type VAD
  • func New() (*VAD, error)
  • func (v *VAD) Process(fs int, audioFrame []byte) (activeVoice bool, err error)
  • func (v *VAD) SetMode(mode int) error
  • func (v *VAD) ValidRateAndFrameLength(rate int, frameLength int) bool

Types

type VAD

type VAD struct {
	// contains filtered or unexported fields
}

func New

func New() (*VAD, error)

func (*VAD) Process

func (v *VAD) Process(fs int, audioFrame []byte) (activeVoice bool, err error)

func (*VAD) SetMode

func (v *VAD) SetMode(mode int) error

func (*VAD) ValidRateAndFrameLength

func (v *VAD) ValidRateAndFrameLength(rate int, frameLength int) bool