一种用于无监督音频恢复的新型U-Net先验_Python_下载.zip

# On the Design of Deep Priors for Unsupervised Audio Restoration Refer to diagram below for a summary of the approach.  ## Requirements * python 3.6 * numpy >= 1.11.0 * pytorch = 1.5.0 * scikit-learn >= 0.18 * matplotlib >= 2.1.0 * librosa ## Denoising Example  ### Perform Denoising Run ```./denoising/script.py``` to perform audio denoising on a single audio excerpt (LJSpeech, Digits, Piano).