EVA-GAN: Enhanced Various Audio Generation via Scalable Generative Adversarial Networks
Open paper
Benchmark
Model
Rank
Results
speech-synthesis-on-libritts
EVA-GAN-big
#3
PESQ: 4.3536
M-STFT: 0.7982
Periodicity: 0.0751
V/UV F1: 0.9745
speech-synthesis-on-libritts
EVA-GAN-base
#8
PESQ: 4.0330
M-STFT: 0.9485
Periodicity: 0.0942
V/UV F1: 0.9658
Rank counts only results with a code link.