Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Text-to-Speech on Spark-TTS generated audio
Loading...
0.2057
FAD (VGGish)
None
0.199904
0.239027
0.27815
0.317273
May 25, 2026
FAD (VGGish)
FAD (CLAP)
MOS (NISQA)
MOS (DNSMOS)
WER
CER
Updated 7d ago
Evaluation Results
Method
Method
Links
FAD (VGGish)
FAD (CLAP)
MOS (NISQA)
MOS (DNSMOS)
WER
CER
None
Watermarking=None
2026.05
0.2057
0.0401
3.37
2.94
0.97
0.18
Base
Watermarking=Base
2026.05
0.2221
0.0484
3.3
2.97
0.98
0.12
Ours
Watermarking=Ours
2026.05
0.3506
0.0472
3.46
2.96
0.99
0.24
Feedback
Search any
task
Search any
task
