CAM++ speaker embedding (zh-cn, 200k speakers), ONNX
An unmodified copy of 3dspeaker_speech_campplus_sv_zh-cn_16k-common.onnx from the
sherpa-onnx speaker-recongition-models release,
mirrored here so that mac-toolbox can download it from a pinned revision.
- Original model:
iic/speech_campplus_sv_zh-cn_16k-common(3D-Speaker, Alibaba Tongyi Lab), Apache-2.0. - Trained on a Mandarin dataset of about 200k speakers.
- Input: 16 kHz mono audio. Output: a 192-dimensional speaker embedding.
- ONNX export: k2-fsa/sherpa-onnx.
- SHA-256:
f682b514c05d947ee3fa91cd6ec6c5c7543479a128373fa29b1faedccd21fd11(matches the sherpa-onnx releasechecksum.txt).
mac-toolbox 的聲紋比對用這個模型;檔案沒有改過,授權照原模型(Apache-2.0)。