Crate wavekat_vad 

WaveKat VAD — Unified voice activity detection with multiple backends.

This crate provides a common VoiceActivityDetector trait with implementations for different VAD backends, enabling experimentation and benchmarking across technologies.

Backends

Backend	Feature	Description
WebRTC	webrtc (default)	Google’s WebRTC VAD — fast, binary output
Silero	silero	Neural network via ONNX Runtime — higher accuracy, continuous probability
TEN-VAD	ten-vad	Agora’s TEN-VAD via ONNX — pure Rust, no C dependency

Feature flags

• webrtc (default) — WebRTC VAD backend
• silero — Silero VAD backend (downloads ONNX model at build time)
• ten-vad — TEN-VAD backend (downloads ONNX model at build time)
• denoise — RNNoise-based noise suppression in the preprocessing pipeline
• serde — Serialize/Deserialize impls for config types

TEN-VAD model license

The TEN-VAD ONNX model is licensed under Apache-2.0 with a non-compete clause by the TEN-framework / Agora. It restricts deployment that competes with Agora’s offerings. Review the TEN-VAD license before using in production.

Example

use wavekat_vad::VoiceActivityDetector;
use wavekat_vad::backends::webrtc::{WebRtcVad, WebRtcVadMode};

let mut vad = WebRtcVad::new(16000, WebRtcVadMode::Quality).unwrap();
let samples = vec![0i16; 160]; // 10ms at 16kHz
let probability = vad.process(&samples, 16000).unwrap();
println!("Speech probability: {probability}");

Re-exports

pub use adapter::FrameAdapter;

pub use error::VadError;

Modules

adapter

Frame adapter for matching audio frames to VAD backend requirements.

backends

error

frame

preprocessing

Audio preprocessing pipeline for improving VAD accuracy.

Structs

VadCapabilities

Describes the audio requirements of a VAD backend.

Traits

VoiceActivityDetector

Common interface for voice activity detection backends.