Target Speaker Extraction on Libri2Mix Noisy
2.21PESQMeanFlow-TSE
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| MeanFlow-TSEType=G2025.12 | 2.21 | 0.82 | 12.85 | 3.17 | 3.55 | 0.73 | |
| AD-FlowTSE2025.12 | 2.15 | 0.81 | 12.69 | 3.11 | 3.48 | 0.87 | |
| SoloSpeech2025.12 | 1.89 | 0.78 | 11.12 | — | 3.76 | — | |
| FlowTSE2025.12 | 1.86 | 0.75 | — | 3.3 | 3.82 | 0.83 | |
| DDTSE2025.12 | 1.6 | 0.71 | — | 3.28 | 3.74 | 0.71 | |
| DiffSep+SVType=G2025.12 | 1.32 | 0.6 | — | 2.78 | 3.63 | 0.62 | |
| Mixture2025.12 | 1.08 | 0.4 | -1.93 | 1.63 | 2.71 | 0.46 |