List your company

Speech Synthesis on LibriTTS

4.5PESQGround Truth
2.7461443.2014723.65684.112128Dec 3, 2025Dec 19, 2025Jan 5, 2026Jan 21, 2026Feb 7, 2026Feb 23, 2026Mar 12, 2026

Evaluation results

MethodLinks
Ground Truth2025.124.50————————
PeriodWaveNFE=16, #Param. (M)=29.81, #MACs (Giga/5s)=4986.562026.034.24———3.6524.749————
RNDVoC-sharedNFE=1, #Param. (M)=3.14, #MACs (Giga/5s)=37.202026.034.226———3.6574.915————
RNDVoC-nonsharedNFE=1, #Param. (M)=9.48, #MACs (Giga/5s)=28.292026.034.085———3.6074.873————
PriorGradNFE=50, #Param. (M)=2.70, #MACs (Giga/5s)=83642026.034.017———3.3294.737————
WaveFMNFE=6, #Param. (M)=19.50, #MACs (Giga/5s)=11882026.033.954———3.1914.943————
ComVo2026.033.8239——4.073.6901—0.84390.090396.090
FreGradNFE=50, #Param. (M)=1.82, #MACs (Giga/5s)=15892026.033.793———3.2174.699————
Vocos2026.033.6266——4.053.6025—0.88560.106195.220.02
BigVGAN2026.033.6122——4.053.5197—0.89940.118194.180.05
HiFiGAN + UnivHDVocoder=HiFiGAN, Discriminator=UnivHD (+H)2025.123.132.8837.85———————
iSTFTNet + UnivHDVocoder=iSTFTNet, Discriminator=UnivHD (+H)2025.123.072.9240.03———————
HiFiGAN + MS-SB-CQTVocoder=HiFiGAN, Discriminator=MS-SB-CQT (+C)2025.123.062.9239.46———————
HiFiGAN + MS-STFTVocoder=HiFiGAN, Discriminator=MS-STFT (+S)2025.123.022.9541.06———————
iSTFTNet + MS-SB-CQTVocoder=iSTFTNet, Discriminator=MS-SB-CQT (+C)2025.123.012.9441.03———————
HiFiGANVocoder=HiFiGAN, Discriminator=Original2025.1232.9946.25———————
iSTFTNet + MS-STFTVocoder=iSTFTNet, Discriminator=MS-STFT (+S)2025.122.992.9640.46———————
iSTFTNetVocoder=iSTFTNet, Discriminator=Original2025.122.952.9643.41———————
HiFi-GAN2026.032.936——43.3453—1.04550.155491.740.09
iSTFTNetImplementation=open-source2026.032.8136——3.983.3591—1.10460.147692.430.04
GT2026.03———4.083.8712————0.14