DiariZen Speaker Segmentation SDK
CPU+NPU hybrid speaker diarization segmentation inference.
Architecture
Audio (16kHz mono, 4s)
→ CPU LayerNorm preprocessing
→ AX650 NPU CNN feature extractor (U16)
→ CPU WavLM Transformer + Conformer + Classifier (FP32, ONNX Runtime)
→ Frame-level log-probabilities (1, 199, 11)
Requirements
- Python 3.8+
- numpy, onnxruntime, soundfile
- pyaxengine (for NPU inference)
Usage
from diarizen_sdk import DiarizenSegmenter
segmenter = DiarizenSegmenter("cnn_features.axmodel", "backend.onnx")
log_probs = segmenter(audio_array, sample_rate=16000)
# log_probs: (1, 199, 11) float32