Silero VAD v4.0 (ONNX)

Silero Voice Activity Detection model v4.0 in ONNX format.

Sourced from snakers4/silero-vad tag v4.0.

Model Details

  • Version: 4.0
  • Architecture: LSTM-based VAD with separate h/c states
  • ONNX Inputs: input, sr, h [2,batch,64], c [2,batch,64]
  • ONNX Outputs: output, hn, cn
  • Size: ~1.8 MB
  • Sample rate: 16kHz (or 8kHz)
  • Window: 512 samples (32ms at 16kHz)
Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support