CycleGAN with Dual Adversarial Loss for Bone-Conducted Speech Enhancement.

Qing Pan Teng Gao Jian Zhou Huabin Wang Liang Tao Hon Keung Kwan

Published in: CoRR (2021)

Keyphrases

speech enhancement
noisy environments
signal to noise ratio
noise reduction
speech signal
wiener filter
single channel
linear prediction
image processing
multiscale
non stationary
speech recognition
vocal tract