跳转至

vad_speechbrain model guide

Overview

vad_speechbrain is a VoiceHub voice activity detection integration. This page is generated from the model registry and its executable data and training contracts, so the documented support stays aligned with code. Open the vad_speechbrain Colab notebook.

Quickstart

python -m pip install voicehub
  1. Install VoiceHub and the provider extra shown above.
  2. Choose a checkpoint that matches this integration.
  3. Place a supported recording at speech.wav.
  4. Run detection and tune the threshold against labeled validation audio.
from voicehub import AutoModelForVoiceActivityDetection

model = AutoModelForVoiceActivityDetection.from_pretrained(
    'speechbrain/vad-crdnn-libriparty',
    model_type='vad_speechbrain',
    device="cpu",
    lazy_load=True,
)
output = model.detect("speech.wav", threshold=0.5)
for segment in output.segments:
    print(segment.start, segment.end, segment.score)

Use only authorized recordings for reference voice, transcription, detection, or evaluation. The example selects a concrete device; verify checkpoint-specific hardware needs and pin an immutable revision before production use.

Supported tasks and capabilities

Property Value
Task Voice activity detection
Architecture speechbrain-crdnn-vad
Runtime VoiceHub-native
Capabilities voice-activity-detection, voicehub-native, safetensors, trusted-checkpoint-conversion, frame-scores, fine-tuning, offline-bidirectional
Reusable components —

Data contract

Property Value
Label boundary Clip-, frame-, or segment-level labels
Required training inputs waveforms, labels

Use authorized audio and preserve annotation provenance. Follow the ASR and VAD data workflow for supported audio forms, timestamp labels, frame targets, leakage-safe splits, and evaluation.

Checkpoints, provenance, and license

Property Value
Default checkpoint speechbrain/vad-crdnn-libriparty
Checkpoint status Registry default; pin an immutable revision for production and reproducible evidence
Implementation voicehub.models.vad_speechbrain.modeling_vad_speechbrain.SpeechBrainVADForVoiceActivityDetection
Configuration voicehub.models.vad_speechbrain.configuration_vad_speechbrain.SpeechBrainVADConfig
Source provenance voicehub/architectures/speechbrain_vad/SOURCE.json
License Checkpoint-specific

No VoiceHub-specific license override is registered. Verify the checkpoint and upstream source terms before use.

The default checkpoint identifies the expected family, not every compatible variant. Confirm the selected checkpoint's revision, access terms, provenance, and license before downloading or redistributing it.

Optimization and training support

All public optimizations enter this model through the shared BaseSpeechModel lifecycle. Use available_optimization_passes() to discover the public pass registry, then apply, inspect, serialize, or restore a plan through the common model API. Application remains fail-closed when the active runtime or hardware cannot satisfy a pass.

Training contract

Property Value
Support native
Family frame-classification
Recipe single-phase
Default phase voice_activity_detection
Training checkpoint speechbrain/vad-crdnn-libriparty
Native training graph yes
Phase Kind Components Required inputs Loss keys
voice_activity_detection objective model waveforms, labels loss

The integration accepts its declared source or prepared contract directly. Call model.validate_training_support() before constructing a trainer. Follow the shared training workflow for a one-step smoke test, validation, checkpoint resume, optimization, and portable export.

Public API

Purpose Public object
Discover get_model_spec('vad_speechbrain')
Load and run AutoModelForVoiceActivityDetection
Configure SpeechBrainVADConfig
Model implementation SpeechBrainVADForVoiceActivityDetection
Normalized output VADOutput
Training contract get_training_spec('vad_speechbrain')
Optimization lifecycle available_optimization_passes, apply_optimization_plan, optimization_manifest, restore_optimization_plan

Related shared documentation: