We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

eess.AS

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Electrical Engineering and Systems Science > Audio and Speech Processing

Title: Enhancing audio quality for expressive Neural Text-to-Speech

Abstract: Artificial speech synthesis has made a great leap in terms of naturalness as recent Text-to-Speech (TTS) systems are capable of producing speech with similar quality to human recordings. However, not all speaking styles are easy to model: highly expressive voices are still challenging even to recent TTS architectures since there seems to be a trade-off between expressiveness in a generated audio and its signal quality. In this paper, we present a set of techniques that can be leveraged to enhance the signal quality of a highly-expressive voice without the use of additional data. The proposed techniques include: tuning the autoregressive loop's granularity during training; using Generative Adversarial Networks in acoustic modelling; and the use of Variational Auto-Encoders in both the acoustic model and the neural vocoder. We show that, when combined, these techniques greatly closed the gap in perceived naturalness between the baseline system and recordings by 39% in terms of MUSHRA scores for an expressive celebrity voice.
Comments: 6 pages, 4 figures, 2 tables, SSW 2021
Subjects: Audio and Speech Processing (eess.AS); Artificial Intelligence (cs.AI)
Cite as: arXiv:2108.06270 [eess.AS]
  (or arXiv:2108.06270v1 [eess.AS] for this version)

Submission history

From: Abdelhamid Ezzerg [view email]
[v1] Fri, 13 Aug 2021 14:32:39 GMT (3214kb,D)

Link back to: arXiv, form interface, contact.