Speech Continuation on Librispeech (test-clean)
0d MAE (min)ground truth
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| ground truthInput=n/a, Output=n/a2021.09 | 0 | 0.463 | 1.32 | 0 | 0.52 | 0.163 | 1 | |
| resynthesizedQuantized=true2021.09 | 0 | 0.464 | 1.32 | 0 | 0.315 | 0.145 | 0.943 | |
| MS-TLMID=12, Input=(u, d, lf), Output=(u, d, lf), Quantized=true, Backbone=Large, Units=HUBERT, Training Dataset=LL6k, Delta=12021.09 | 0.536 | 0.242 | 0.946 | 0.077 | 0.324 | 0.149 | 0.499 | |
| MS-TLMID=11, Input=(u, d, lf), Output=(u, d, lf), Quantized=false, Backbone=Large, Units=HUBERT, Training Dataset=LL6k, Delta=12021.09 | 0.539 | 0.344 | 0.94 | 0.081 | 0.494 | 0.076 | 0.498 | |
| MS-TLMID=9, Input=u, Output=(u, d, lf), Quantized=false, Backbone=Large, Units=HUBERT, Training Dataset=LL6k, Delta=12021.09 | 0.542 | 0.176 | 0.942 | 0.084 | 0.093 | 0.081 | 0.488 | |
| MS-TLMID=10, Input=u, Output=(u, d, lf), Quantized=true, Backbone=Large, Units=HUBERT, Training Dataset=LL6k, Delta=12021.09 | 0.542 | 0.086 | 0.965 | 0.096 | 0.217 | 0.147 | 0.489 |