FlaxSpeaker: ls_conformer_small_ext_softmax (conformer - small)

Disclaimer & Open-Source Reproduction Notice

This is not an officially supported Google product.

This model checkpoint and the accompanying FlaxSpeaker library are an independent, clean-room open-source reproduction in JAX/Flax of speaker recognition algorithms published in:

• GE2E: Generalized End-to-End Loss for Speaker Verification (arXiv:1710.10467)
• Dr-Vectors: Decision Residual Networks and an Improved Loss for Speaker Recognition (arXiv:2104.01989)
• Conformer-ATP: Attentive Temporal Pooling for Conformer-based Streaming Language Identification in Long-form Speech (arXiv:2202.12163)
• PFAS: Parameter-Free Attentive Scoring for Speaker Verification (arXiv:2203.05642v3)

The experiments in those original papers were conducted using Google's proprietary internal infrastructure and internal datasets, not this open-source library or public dataset split.

Model Summary

Property	Value
Model ID	ls_conformer_small_ext_softmax
Hugging Face Repository	wq2012/ls_conformer_small_ext_softmax
Hugging Face Collection	FlaxSpeaker Collection
GitHub Repository	wq2012/FlaxSpeaker
PyPI Package	flaxspeaker
Backbone Architecture	conformer (small)
Temporal Pooling	atp
Training Loss	extended_set_softmax
Verification Scoring	cosine
Output Embedding Dim	192
Encoder Parameters	2,029,894
Training Domain	librispeech
Audio Frontend	80-dim log_mel @ 16000 Hz + Energy VAD + CMVN

Model Files Included

• model.safetensors: Primary model weights in SafeTensors format
• flax_model.msgpack: Native Flax Msgpack serialized parameters
• config.json: FlaxSpeakerConfig model & training architecture configuration
• preprocessor_config.json: WhisperFeatureExtractor-compatible frontend configuration
• eval_results.json: Verified multi-domain evaluation metrics (1-utt & 3-utt EER, minDCF, ROC-AUC, 95% bootstrap CIs)

Quickstart Usage

Python API (from Hugging Face Hub or Local Directory)

from flaxspeaker import load_pretrained

model = load_pretrained("wq2012/ls_conformer_small_ext_softmax")
emb = model.extract_embedding("path/to/utterance.flac")
score = model.verify("path/to/enroll.flac", "path/to/test.flac")
print(f"Embedding shape: {emb.shape}, Verification score: {score:.4f}")

Command-Line Inference

python scripts/inference.py --model_dir wq2012/ls_conformer_small_ext_softmax \
    --enroll_audio path/to/enroll.flac --test_audio path/to/test.flac

Multi-Domain Evaluation Results

1-Utterance Verification Across Evaluation Domains

Evaluation Split	Trials	EER (%)	95% Bootstrap CI (%)	minDCF@0.01	minDCF@0.10	ROC-AUC
LibriSpeech (English Read)	3,000	10.03%	[9.10, 11.12]	0.7967	0.5493	0.9604
CN-Celeb2 (Chinese Multi-Genre)	3,000	29.87%	[28.03, 31.40]	0.9593	0.9233	0.7851
VoxCeleb1 (In-the-Wild Celebrity)	2,000	24.10%	[22.35, 25.95]	0.9490	0.8890	0.8446
Universal Pooled (All 3 Domains)	8,000	21.05%	[20.11, 21.93]	0.9555	0.8115	0.8779
Universal Macro-Average (3 Domains)	8,000	21.33%	—	0.9017	0.7872	0.8634

3-Utterance Multi-Enrollment Verification Across Evaluation Domains

Evaluation Split	Trials	3-Utt EER (%)	95% Bootstrap CI (%)	minDCF@0.01	ROC-AUC
LibriSpeech (3-Utt)	1,000	6.20%	[4.65, 7.55]	0.4980	0.9846
CN-Celeb2 (3-Utt)	1,000	25.50%	[22.95, 28.50]	0.9120	0.8314
VoxCeleb1 (3-Utt)	1,000	20.20%	[18.00, 22.75]	0.9420	0.8820
Universal Pooled (3-Utt)	3,000	17.53%	[16.20, 18.73]	0.8853	0.9114
Universal Macro-Average (3-Utt)	3,000	17.30%	—	0.7840	0.8993

Citation

@inproceedings{wan2018generalized,
  title={Generalized End-to-End Loss for Speaker Verification},
  author={Wan, Li and Wang, Quan and Papir, Alan and Moreno, Ignacio Lopez},
  booktitle={IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
  pages={4879--4883},
  year={2018}
}

@inproceedings{pelecanos2021dr,
  title={Dr-Vectors: Decision Residual Networks and an Improved Loss for Speaker Recognition},
  author={Pelecanos, Jason and Wang, Quan and Moreno, Ignacio Lopez},
  booktitle={Interspeech},
  pages={4603--4607},
  year={2021}
}

@inproceedings{wang2022attentive,
  title={Attentive Temporal Pooling for Conformer-based Streaming Language Identification in Long-form Speech},
  author={Wang, Quan and Yu, Yang and Pelecanos, Jason and Huang, Yiling and Moreno, Ignacio Lopez},
  booktitle={The Speaker and Language Recognition Workshop (Odyssey)},
  year={2022}
}

@inproceedings{pelecanos2022parameter,
  title={Parameter-Free Attentive Scoring for Speaker Verification},
  author={Pelecanos, Jason and Wang, Quan and Huang, Yiling and Moreno, Ignacio Lopez},
  booktitle={The Speaker and Language Recognition Workshop (Odyssey)},
  year={2022}
}