Skip to main content
Version: 0.10.0

Type Alias: FsmnVoiceActivityDetector

FsmnVoiceActivityDetector = object

Defined in: extensions/speech/tasks/fsmnVoiceActivityDetection.ts:180

Voice activity detection task runner using the FSMN-VAD model.

Properties

detectVoice()

readonly detectVoice: (waveform, options?) => Promise<VadSegment[]>

Defined in: extensions/speech/tasks/fsmnVoiceActivityDetection.ts:196

Asynchronously detects speech segments within a mono waveform sampled at FSMN_VAD_SAMPLE_RATE_HZ.

Parameters

waveform

Float32Array

The input audio samples (16 kHz mono Float32 PCM).

options?

VadOptions

Optional per-call overrides of the detection thresholds. See VadOptions.

Returns

Promise<VadSegment[]>

A promise resolving to the detected speech segments, in seconds.

Throws

With code RESOURCE_BUSY if the model is in use, or RESOURCE_DISPOSED if disposed.


detectVoiceOnStream()

readonly detectVoiceOnStream: (chunk, options?) => VadEvent | undefined

Defined in: extensions/speech/tasks/fsmnVoiceActivityDetection.ts:217

Appends a live audio chunk to a bounded rolling window, runs detection over that window and reports a VadEvent when speech starts or stops, otherwise undefined. Designed to be driven straight from a recorder callback. Detection runs synchronously on the calling thread (a few ms).

Parameters

chunk

Float32Array

The newly captured audio samples (16 kHz mono Float32 PCM).

options?

VadStreamOptions

Optional overrides of the detection thresholds and margin. See VadStreamOptions.

Returns

VadEvent | undefined

'speechStart' or 'speechEnd' if state transitioned, otherwise undefined.

Throws

With code RESOURCE_BUSY if the model is in use, or RESOURCE_DISPOSED if disposed.


detectVoiceWorklet()

readonly detectVoiceWorklet: (waveform, options?) => VadSegment[]

Defined in: extensions/speech/tasks/fsmnVoiceActivityDetection.ts:202

Synchronous version of detectVoice to be executed directly on the caller or worklet thread.

Parameters

waveform

Float32Array

options?

VadOptions

Returns

VadSegment[]


dispose()

readonly dispose: () => void

Defined in: extensions/speech/tasks/fsmnVoiceActivityDetection.ts:184

Releases all allocated native resources.

Returns

void


resetStream()

readonly resetStream: () => void

Defined in: extensions/speech/tasks/fsmnVoiceActivityDetection.ts:225

Clears the rolling window and speaking state used by detectVoiceOnStream.

Returns

void