Downloads · 30 days
0
silero-vad-ov
jinhwan000/silero-vad-ov
silero-vad-ov is a machine learning model from jinhwan000. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as mit.
OpenVINO IR conversion of silero-vad v5 (MIT), for 16 kHz streaming inference.
Access
Public
Updated Jun 12, 2026
Repo size
1.2 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin1.2 MB · 89%
At a glance
- License
- mit
- Access
- Public
- Created
- Jun 12, 2026
- Updated
- Jun 12, 2026
- SHA
- d795f76c
Model card
From the Hugging Face model README
silero-vad-ov
OpenVINO IR conversion of silero-vad v5 (MIT), for 16 kHz streaming inference.
- The 8 kHz branch of the original ONNX graph is removed (the If node then-branch is inlined and the sr input folded to 16000) for OpenVINO ONNX frontend compatibility.
- Input input [1, 576] = 64-sample context (tail of the previous chunk) + 512 new samples, with state [2, 1, 128] — matching the silero-vad OnnxWrapper streaming convention. Outputs: output [1, 1] speech probability and stateN.
- Parity vs onnxruntime on real audio (200 consecutive windows): max probability diff 5.96e-7, zero threshold-crossing (0.5) mismatches.
- License
- mit
- Created
- Jun 12, 2026
- Updated
- Jun 12, 2026