CycleGAN with Dual Adversarial Loss for Bone-Conducted Speech Enhancement
Qing Jun Pan, Yan Pan, Jian Zhou, Huabin Wang, Liang Tao, Hon Keung Kwan · TENCON 2022 - 2022 IEEE Region 10 Conference (TENCON) · 2022
Compared with air-conducted speech, bone-conducted speech has the unique advantage of shielding background noise. Enhancement of bone-conducted speech helps to improve its quality and intelligibility. In this paper, a novel CycleGAN with dual adversarial loss (CycleGAN-DAL) is proposed for bone-conducted speech enhancement. The proposed method uses the adversarial loss and the cycle-consistent loss simultaneously to learn forward and cyclic mapping, in which the adversarial loss is replaced with the classification adversarial loss and the defect adversarial loss to consolidate the forward mapping. Compared with conventional baseline methods, it can learn feature mapping between bone-conducted speech and target speech without additional air-conducted speech assistance. Moreover, the proposed method also avoids the over-smooth problem which is occurred commonly in conventional statistical-based models. Experimental results show that the proposed method outperforms baseline methods such as CycleGAN, GMM, and BLSTM.