Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Video-to-speech synthesis on LRS2-BBC (test)
Loading...
3.921
UTMOS
Ours
1.28564
1.96982
2.654
3.33818
Mar 21, 2025
UTMOS
DNSMOS
RMSE f0
WER
Updated 4d ago
Evaluation Results
Method
Method
Links
UTMOS
DNSMOS
RMSE f0
WER
Ours
Speaker embedding type...
2025.03
3.921
2.586
43.441
39.37
Ours
Speaker embedding type...
2025.03
3.881
2.552
43.702
39.05
Ground Truth
2025.03
3.013
2.256
-
8.93
DiffV2S
Speaker embedding type...
2025.03
2.945
2.363
44.414
54.86
Intelligible
Speaker embedding type...
2025.03
2.331
2
41.233
39.53
LTBS
Speaker embedding type...
2025.03
2.288
2.174
43.653
94.25
SVTS
Speaker embedding type...
2025.03
1.387
1.434
53.475
83.38
Feedback
Search any
task
Search any
task