diaric
Backend-free speaker-diarization core: clustering (offline AHC→VBx, online), PLDA projection, pipeline assembly, reconstruction/RTTM, kaldi-fbank DSP and embedding types, adaptive score normalization (AS-Norm1), plus SIMD/mmap numeric ops. The ONNX/Torch model runners live in the `diarization` crate.
View on Github
Activity
- Latest release
- 1w ago
- Total releases
- 2
- Cadence
- ~2 months
- Last 12 months
- 2
Reach
- Downloads
- 1.0k
- Stars
- 0
Details
- License
- MIT OR Apache-2.0 AND Apache-2.0 AND MIT AND CC-BY-4.0 AND BSD-2-Clause AND BSD-3-Clause
- First release
- Jul 17, 2026
Releases