feat(audio): add Apple CoreML Silero VAD

This commit is contained in:
Edison Jwa
2026-06-02 01:52:08 +09:00
parent 71b7502ab7
commit 813a38e92b
6 changed files with 411 additions and 23 deletions
+42 -5
View File
@@ -116,6 +116,7 @@ mod inner {
mic_gain: f32,
voice_activity_selector: Option<Arc<crate::TransmitModeSelector>>,
vad_detector: crate::vad::WebRtcFallbackVad,
silero_coreml_worker: Option<crate::vad::apple_coreml::AppleCoreMlVadWorker>,
current_vad_backend: crate::VadBackend,
capture_frame_seq: u64,
vad_state: crate::voice_activity::VoiceActivityStateMachine,
@@ -152,6 +153,7 @@ mod inner {
mic_gain: params.mic_gain,
voice_activity_selector: params.voice_activity_selector.clone(),
vad_detector: crate::vad::WebRtcFallbackVad::default(),
silero_coreml_worker: None,
current_vad_backend: crate::VadBackend::WebrtcVad,
capture_frame_seq: 0,
vad_state: crate::voice_activity::VoiceActivityStateMachine::default(),
@@ -272,6 +274,7 @@ mod inner {
.map(|selector| selector.mode() == crate::TransmitMode::VoiceActivity)
.unwrap_or(false);
if !voice_activity_mode {
self.silero_coreml_worker = None;
self.current_vad_backend = crate::VadBackend::Disabled;
self.fallback_warned_backend = None;
self.audio_processing_stats.set_vad_fallback_active(false);
@@ -297,9 +300,16 @@ mod inner {
self.current_vad_backend = vad_backend;
self.fallback_warned_backend = None;
if vad_backend == crate::VadBackend::SileroOnnx {
self.mark_vad_fallback_active(crate::VadBackend::SileroOnnx);
self.audio_processing_stats.set_vad_fallback_active(true);
self.silero_coreml_worker =
crate::vad::apple_coreml::AppleCoreMlVadWorker::try_new();
if self.silero_coreml_worker.is_none() {
self.mark_vad_fallback_active(crate::VadBackend::SileroOnnx);
self.audio_processing_stats.set_vad_fallback_active(true);
} else {
self.audio_processing_stats.set_vad_fallback_active(false);
}
} else {
self.silero_coreml_worker = None;
self.audio_processing_stats.set_vad_fallback_active(false);
}
self.vad_state.reset();
@@ -307,6 +317,7 @@ mod inner {
let (vad_probability, active) = if voice_activity_mode {
self.capture_frame_seq = self.capture_frame_seq.wrapping_add(1);
let capture_seq = self.capture_frame_seq;
let mut used_fallback_vad = false;
let vad = if vad_backend == crate::VadBackend::Disabled {
crate::vad::VadOutput {
@@ -314,9 +325,35 @@ mod inner {
speech: true,
}
} else if vad_backend == crate::VadBackend::SileroOnnx {
used_fallback_vad = true;
self.mark_vad_fallback_active(vad_backend);
crate::vad::VoiceActivityDetector::process_10ms(&mut self.vad_detector, &frame)
if let Some(worker) = self.silero_coreml_worker.as_ref() {
let enqueued = worker.try_send(capture_seq, &frame);
if !worker.is_stale(capture_seq) {
let p = worker.latest_probability();
crate::vad::VadOutput {
probability: p,
speech: p >= 0.5,
}
} else if enqueued {
crate::vad::VadOutput {
probability: 0.0,
speech: false,
}
} else {
used_fallback_vad = true;
self.mark_vad_fallback_active(vad_backend);
crate::vad::VoiceActivityDetector::process_10ms(
&mut self.vad_detector,
&frame,
)
}
} else {
used_fallback_vad = true;
self.mark_vad_fallback_active(vad_backend);
crate::vad::VoiceActivityDetector::process_10ms(
&mut self.vad_detector,
&frame,
)
}
} else {
crate::vad::VoiceActivityDetector::process_10ms(&mut self.vad_detector, &frame)
};