Emilia
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Speech Coding | Emilia clean subset (test) | DNSMOS3.646 | 10 | |
| Text-to-Speech | Emilia EN speaking-rate | MUSHRA Score68.8 | 9 | |
| Text-to-Speech | Emilia | Word Error Rate (WER)2.2 | 6 | |
| Speaker Unlearning | Emilia (Seen opt-out set (-SO)) | WER3.13 | 5 | |
| Semantic Evaluation | Emilia (val) | PPL (Initial)8.42 | 4 | |
| Speech Reconstruction | EMILIA KO | ViSQOL4.281 | 2 | |
| Speech Reconstruction | EMILIA JA | ViSQOL4.335 | 2 | |
| Speech Reconstruction | EMILIA FR | ViSQOL4.313 | 2 | |
| Speech Reconstruction | EMILIA DE | ViSQOL4.382 | 2 |
Showing 9 of 9 rows