Multi-speaker Dialogue Synthesis on CoVoMix2 Dialogue-WildRef (test)
0.156cpWERMOSS-TTSD
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| MOSS-TTSD2026.06 | 0.156 | 0.39 | 84.4 | 0.059 | 0.295 | 3.45 | 4.21 | |
| SCENA2026.06 | 0.167 | 0.424 | 81.9 | 0.022 | 0.348 | 3.3 | 4.28 | |
| VibeVoicemodel_size=7B2026.06 | 0.172 | 0.386 | 84.1 | 0.045 | 0.317 | 2.56 | 2.91 | |
| ZipVoice-Dialog2026.06 | 0.173 | 0.396 | 82.5 | 0.038 | 0.315 | 3.2 | 4.19 | |
| VibeVoicemodel_size=1.5B2026.06 | 0.202 | 0.365 | 82.6 | 0.089 | 0.293 | 2.33 | 2.85 | |
| Diasource=Nari Labs2026.06 | 0.272 | 0.278 | 75.2 | 0.086 | 0.256 | 2.45 | 3.92 |