Target Speaker Extraction on REAL-T CHiME-6 English subset
1.896DNSMOS OVRLMeanFlowTSE
Evaluation Results
| Method | Links | |
|---|---|---|
| MeanFlowTSEMR predictor=with, Training set=Libri2Mix noisy, Zero-shot=true2026.03 | 1.896 | |
| AlphaFlowTSEMR predictor=with, Training set=Libri2Mix noisy, Zero-shot=true2026.03 | 1.858 | |
| AlphaFlowTSEMR predictor=without, Training set=Libri2Mix noisy, Zero-shot=true2026.03 | 1.843 | |
| AD-FlowTSEMR predictor=with, Training set=Libri2Mix noisy, Zero-shot=true2026.03 | 1.597 | |
| AD-FlowTSEMR predictor=without, Training set=Libri2Mix noisy, Zero-shot=true2026.03 | 1.46 | |
| MeanFlowTSEMR predictor=without, Training set=Libri2Mix noisy, Zero-shot=true2026.03 | 1.174 |