pepper.framework.sensor.vad module

class pepper.framework.sensor.vad.VAD(microphone)

Bases: object

Perform Voice Activity Detection on Microphone Input

Parameters:microphone (AbstractMicrophone) –

AUDIO_FRAME_MS = 10

AUDIO_TYPE

alias of numpy.int16

AUDIO_TYPE_BYTES = 2

BUFFER_SIZE = 100

MODE = 3

VOICE_THRESHOLD = 0.6

VOICE_WINDOW = 50

activation

VAD Activation

Returns:activation Return type:float

microphone

VAD Microphone

Returns:microphone Return type:AbstractMicrophone

voices

Get Voices from Microphone Stream

Yields:voices (Iterable[Voice])

class pepper.framework.sensor.vad.Voice

Bases: object

Voice Object (for Voice Activity Detection: VAD)

add_frame(frame)

Add Voice Frame (done by VAD)

Parameters:frame (np.ndarray) –

audio

Get Voice Audio (Concatenated Frames)

Returns:audio Return type:np.ndarray

frames

Get Voice Frames (chunks of audio)

Yields:frames (Iterable of np.ndarray)