Struct LocalVadDetector

pub struct LocalVadDetector { /* private fields */ }

Local voice activity detector.

Provides voice activity detection using local processing without external API calls. Uses the voice_activity_detector_silero_v5 crate for speech detection and analysis.

Implementations

impl LocalVadDetector

pub fn new(config: VadConfig) -> Result<Self>

Create a new local VAD detector.

Arguments

• config - VAD configuration parameters

Returns

A new LocalVadDetector instance

Errors

Returns an error if the audio processor cannot be initialized

pub async fn detect_speech_from_data( &self, audio_data: ProcessedAudioData, ) -> Result<VadResult>

Detect speech activity in a ProcessedAudioData (for partial audio analysis).

Arguments

• audio_data - Pre-processed audio data (can be cropped)

Returns

VAD analysis results including speech segments and metadata

pub fn calculate_chunk_size(&self, sample_rate: u32) -> usize

Dynamically calculates the optimal VAD chunk size for a given audio sample rate using the Silero VAD V5 model.

This function selects a chunk size (in samples) compatible with the Silero VAD V5 model’s strict requirements. For 8 kHz audio, only a 256-sample window is supported. For 16 kHz audio, only a 512-sample window is supported. For sample rates that are a multiple of 16 kHz (e.g., 32 kHz, 48 kHz), a 512-sample window is also used, as required by the model.

Arguments

• sample_rate: The audio sample rate in Hz (e.g., 8000, 16000)

Returns

The chunk size in number of samples, as required by the Silero VAD V5 model.

Model Reference

This logic follows the requirements of the Silero VAD V5 model.

Panics

This function will panic if the sample rate is not supported by the model.

Examples

use subx_cli::services::vad::LocalVadDetector;
let detector = LocalVadDetector::new(Default::default()).unwrap();
assert_eq!(detector.calculate_chunk_size(8000), 256);
assert_eq!(detector.calculate_chunk_size(16000), 512);

pub fn audio_processor(&self) -> &VadAudioProcessor

Get the internal VadAudioProcessor instance (for advanced use, e.g. partial audio cropping)