List your company

Text-to-Speech on LibriTTS (test)

4.54MOSGT
3.14643.50823.874.2318May 22, 2020Nov 19, 2020May 20, 2021Nov 18, 2021May 19, 2022Nov 17, 2022May 18, 2023

Evaluation results

MethodLinks
GT2020.054.54————
GT2023.054.400——
GT(voc.)2023.054.2600——
GTSynthesis Pipeline=Mel + WaveGlow2020.054.22————
VB + CLAPSpeechnumber of parameters=41.54M2023.054.0610.9350.89——
VB + BERTnumber of parameters=132.69M2023.053.9511.5151.27——
VB + A³Tnumber of parameters=59.73M2023.053.8211.7151.98——
VBnumber of parameters=23.02M2023.053.8111.9652.51——
PB + CLAPSpeechnumber of parameters=30.51M2023.053.7113.3726.46——
PB + BERTnumber of parameters=109.48M2023.053.613.8226.7——
PB + A³Tnumber of parameters=48.25M2023.053.5413.6727.03——
Glow-TTSTemperature (T)=0.667, Synthesis Pipeline=Mel + WaveGlow2020.053.45————
PBnumber of parameters=11.99M2023.053.4314.2627.42——
Tacotron2Synthesis Pipeline=Mel + WaveGlow2020.053.35————
Glow-TTSTemperature (T)=0.500, Synthesis Pipeline=Mel + WaveGlow2020.053.31————
Glow-TTSTemperature (T)=0.333, Synthesis Pipeline=Mel + WaveGlow2020.053.2————
CosyVoice2NFEs=102026.05———5.230.9227
CosyVoice3NFEs=102026.05———5.980.9152
Ground Truth2026.05———2.210.9218
ImmersiveTTSNFEs=252026.05———9.890.8859
VoiceDiTNFEs=2002026.05———11.080.8942
VoiceLDMNFEs=2002026.05———14.010.7601