Generative Models for Improved Naturalness, Intelligibility, and Voicing of Whispered Speech

Autor:	Wagner, Dominik, Bayerl, Sebastian P., Maruri, Hector A. Cordourier, Bocklet, Tobias
Rok vydání:	2022
Předmět:	Computer Science - Sound Electrical Engineering and Systems Science - Audio and Speech Processing
Druh dokumentu:	Working Paper
DOI:	10.1109/SLT54892.2023.10022796
Popis:	This work adapts two recent architectures of generative models and evaluates their effectiveness for the conversion of whispered speech to normal speech. We incorporate the normal target speech into the training criterion of vector-quantized variational autoencoders (VQ-VAEs) and MelGANs, thereby conditioning the systems to recover voiced speech from whispered inputs. Objective and subjective quality measures indicate that both VQ-VAEs and MelGANs can be modified to perform the conversion task. We find that the proposed approaches significantly improve the Mel cepstral distortion (MCD) metric by at least 25% relative to a DiscoGAN baseline. Subjective listening tests suggest that the MelGAN-based system significantly improves naturalness, intelligibility, and voicing compared to the whispered input speech. A novel evaluation measure based on differences between latent speech representations also indicates that our MelGAN-based approach yields improvements relative to the baseline. Comment: Accepted at SLT 2022
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2212.01775 Zobrazit plný text záznamu View this record from Arxiv