TOLD

Reported on 5 benchmarks across 1 task · 1 paper · 2 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Speech5 results

Speaker DiarizationonCALLHOME
CF· 2023-03-08
2.94
SOTA
TOLD: A Novel Two-Stage Overlap-Aware Framework for Speaker Diarization arXiv:2303.05397
Speaker DiarizationonCALLHOME
MI· 2023-03-08
4.8
SOTA
TOLD: A Novel Two-Stage Overlap-Aware Framework for Speaker Diarization arXiv:2303.05397
Speaker DiarizationonCALLHOME
DER(%)· 2023-03-08
10.14
best: 24.05 (COS+NJW-SC (Oracle SAD))
TOLD: A Novel Two-Stage Overlap-Aware Framework for Speaker Diarization arXiv:2303.05397
Speaker DiarizationonCALLHOME
DER(ig olp)· 2023-03-08
7.37
best: 9.14 (EEND-OLA)
TOLD: A Novel Two-Stage Overlap-Aware Framework for Speaker Diarization arXiv:2303.05397
Speaker DiarizationonCALLHOME
FA· 2023-03-08
2.4
best: 7.42 (SA-EEND (2-spk, no-adapt))
TOLD: A Novel Two-Stage Overlap-Aware Framework for Speaker Diarization arXiv:2303.05397