import numpy as np
import librosa
#Read melpectogram of npy file
mel = np.load("melspe.npy")
#mel -> audio
audio = librosa.feature.inverse.mel_to_audio(mel)
#Export wav
librosa.output.write_wav("audio.wav",audio,22050)
I had a chance to use it, but I couldn't find an article, so I wrote it. I hope it's okay ... not L ... please do something about it
Recommended Posts