BigVGAN
🔊
108
Generate high-quality audio from input audio
BigVGAN
BigVGAN is a universal neural vocoder that generates audio waveform using mel spectrogram as input.
-
-
nvidia/bigvgan_v2_44khz_128band_512x
Audio-to-Audio • Updated • 293k • 63 -
nvidia/bigvgan_v2_44khz_128band_256x
Audio-to-Audio • Updated • 231 • 7 -
nvidia/bigvgan_v2_24khz_100band_256x
Audio-to-Audio • Updated • 6.6k • 18 -
nvidia/bigvgan_v2_22khz_80band_256x
Audio-to-Audio • Updated • 3.93M • 24 -
nvidia/bigvgan_v2_22khz_80band_fmax8k_256x
Audio-to-Audio • Updated • 1.36k • 2 -
nvidia/bigvgan_24khz_100band
Audio-to-Audio • Updated • 304 • 3 -
nvidia/bigvgan_base_24khz_100band
Audio-to-Audio • Updated • 172 -
nvidia/bigvgan_22khz_80band
Audio-to-Audio • Updated • 17.8k • 1 -
nvidia/bigvgan_base_22khz_80band
Audio-to-Audio • Updated • 943 -
BigVGAN: A Universal Neural Vocoder with Large-Scale Training
Paper • 2206.04658 • Published • 4