Demo for LibriTTS Testset

Audio Reconstruction from Mel-spectrogram using different models.

Ground Truth
Vocos
BigVGAN
RFWave