Parametric Resynthesis With Neural Vocoders

Abstract

Noise suppression systems generally produce output speech with copromised quality. We propose to utilize the high quality speech generation capability of neural vocoders for noise suppression. We use a neural network to predict clean mel-spectrogram features from noisy speech and then compare two neural vocoders, WaveNet and WaveGlow, for synthesizing clean speech from the predicted mel spectrogram. Both WaveNet and WaveGlow achieve better subjective and objective quality scores than the source separation model Chimera++. Further, WaveNet and WaveGlow also achieve significantly better subjective quality ratings than the oracle Wiener mask. Moreover, we observe that between WaveNet and WaveGlow, WaveNet achieves the best subjective quality scores, although at the cost of much slower waveform generation.

An Overview of Voice Conversion and Its Challenges: From Statistical Modeling to Deep Learning

“…WaveGlow benefits from the best of Glow and WaveNet so as to provide fast, efficient and high-quality audio synthesis, without the need for auto-regression. We note that WaveGlow is implemented using only a single network with a single cost function, that is to maximize the likelihood of the training data, which makes the training procedure simple and stable [105].…”

Deep Griffin–Lim Iteration: Trainable Iterative Phase Reconstruction Using Neural Network

“…In the subjective test, the official 6 The audio samples of WN were generated by using Yamamoto's open source implementation: https://doi.org/10.5281/zenodo.1472609. This code has been utilized as a reference in several papers [67], [75], [76]. 7 The audio samples were brought from Google Drive: https://drive.google.…”

Speaker Independence of Neural Vocoders and Their Effect on Parametric Resynthesis Speech Enhancement

“…Speech samples are 8-bit mu-law quantized. We use the officially published LPCNet implementation 2 with 640 units in GRU-A and 16 units in GRU-B. We refer to the PR system with LPCNet as its vocoder as PR-LPCNet.…”

“…Parametric Resynthesis (PR) systems [2,3] predict clean acoustic parameters from noisy speech and synthesize speech from these predicted parameters using a speech synthesizer or vocoder. Current speech synthesizers are trained to generate high quality speech for a single speaker.…”

“…Current speech synthesizers are trained to generate high quality speech for a single speaker. In previous work we showed that a single speaker PR system can synthesize very high quality clean speech at 22 KHz [2] and performs better than the corresponding TTS system [3]. Hence, a critical question is whether these systems can be generalized to unknown speakers.…”