Vowen offline speaker diarization models (Windows)

segmentation.onnx
  pyannote/segmentation-3.0 (MIT License — see LICENSE-pyannote-segmentation.txt)
  ONNX export by the sherpa-onnx project (k2-fsa).
  https://huggingface.co/pyannote/segmentation-3.0

embedding.onnx
  WeSpeaker en_voxceleb_resnet34_LM speaker embedding model (CC-BY-4.0)
  https://github.com/wenet-e2e/wespeaker
  ONNX export distributed by the sherpa-onnx project (k2-fsa).
