Speech Enhancement on TCD-TIMIT 3 Speakers (test)
5.37SDRAV c-ref
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| AV c-reftype=Speaker-independent, features=Landmark motion vectors, talkers=32018.11 | 5.37 | 2.45 | 2.58 | |
| VL2M_reftype=Speaker-independent, features=Landmark motion vectors, talkers=32018.11 | 5.27 | 2.44 | 2.54 | |
| AV concattype=Speaker-independent, features=Landmark motion vectors, talkers=32018.11 | 5.15 | 2.41 | 2.52 | |
| VL2Mtype=Speaker-independent, features=Landmark motion vectors, talkers=32018.11 | -0.51 | 1.99 | 1.98 | |
| Noisytype=Input mixed-speech signal, talkers=32018.11 | -3.42 | 1.92 | 2.04 |