SUPERB
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Universal Speech Representation Evaluation | SUPERB Benchmark | Overall Score1,010.29 | 60 | |
| Speech Processing | SUPERB | KWS Acc0.988 | 52 | |
| Slot Filling | SUPERB SF | F1 Score92.21 | 33 | |
| Speaker Identification | SUPERB SID | Accuracy81.42 | 14 | |
| Automatic Speech Recognition | SUPERB ASR | WER6.42 | 14 | |
| Phoneme Recognition | SUPERB PR | PER5.41 | 14 | |
| Automatic Speech Recognition | SUPERB 48kHz | WER5.53 | 12 | |
| Automatic Speech Recognition | SUPERB 24kHz | WER6.35 | 12 | |
| Automatic Speech Recognition | SUPERB 22.05kHz | WER2.9 | 12 | |
| Automatic Speech Recognition | SUPERB 16kHz | WER5.89 | 12 | |
| Speech Processing | SUPERB | PER3.21 | 9 | |
| Voice Conversion | SUPERB SG | MCD7.26 | 7 | |
| Speech Separation | SUPERB SG | SI-SDRi11.23 | 7 | |
| Speech Enhancement | SUPERB-SG | PESQ2.18 | 7 | |
| Speech Understanding | SUPERB | PR (PER)3.23 | 7 | |
| Automatic Speech Recognition | SUPERB ASR Librispeech (test) | WER3.4 | 6 | |
| Speech Probing | SUPERB | PR Error Rate7.49 | 4 |
Showing 17 of 17 rows