Views
No views yet
| Artifact | Source |
|---|---|
Segmentation.mlmodelc | pyannote/speaker-diarization-community-1/segmentation/pytorch_model.bin |
FBank.mlmodelc | Deterministic feature-extraction graph configured from the embedding checkpoint |
Embedding.mlmodelc | pyannote/speaker-diarization-community-1/embedding/pytorch_model.bin |
PLDA.mlmodelc, PldaRho.mlmodelc | plda/plda.npz and plda/xvec_transform.npz |
plda-parameters.json, xvector-transform.json | Serialized tensors from the same two PLDA files |
Segmentation, FBank,
Embedding, and PldaRho under mlpackages/; no uncompiled PLDA package was
published. These are Core ML conversions, not fine-tuned models.provenance.json. Their historical
source lineage has been reconstructed from the public Community-1 repository,
artifact metadata, and the published conversion source. The exact local
upstream checkout and Mobius commit used for the original 2025 conversion were
not recorded, so the existing binaries are not described as a fully attested
reproducible build.pyannote/speaker-diarization-community-1@3533c8cf.
The public historical conversion reference is
FluidInference/mobius@33fd6eab.FluidInference/mobius@ffbc3c8.pyannote_segmentation.mlmodelc,
wespeaker.mlmodelc, wespeaker_v2.mlmodelc, and wespeaker_int8.mlmodelc
for FluidAudio's legacy online diarizer. They predate the Community-1 export,
record a different toolchain, and are not included in the Community-1
provenance or license-scope confirmation in NOTICE.md.1@inproceedings{Plaquet23,
2 author={Alexis Plaquet and Hervé Bredin},
3 title={{Powerset multi-class cross entropy loss for neural speaker diarization}},
4 year=2023,
5 booktitle={Proc. INTERSPEECH 2023},
6}1@inproceedings{Wang2023,
2 title={Wespeaker: A research and production oriented speaker embedding learning toolkit},
3 author={Wang, Hongji and Liang, Chengdong and Wang, Shuai and Chen, Zhengyang and Zhang, Binbin and Xiang, Xu and Deng, Yanlei and Qian, Yanmin},
4 booktitle={ICASSP 2023, IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
5 pages={1--5},
6 year={2023},
7 organization={IEEE}
8}1@article{Landini2022,
2 author={Landini, Federico and Profant, J{\'a}n and Diez, Mireia and Burget, Luk{\'a}{\v{s}}},
3 title={{Bayesian HMM clustering of x-vector sequences (VBx) in speaker diarization: theory, implementation and analysis on standard tasks}},
4 year={2022},
5 journal={Computer Speech & Language},
6}