An Open Source HMM-based Text-to-Speech System for Brazilian Portuguese

In this paper we present different resources for the study of spoken Brazilian Portuguese, developed within the C-ORAL-BRASIL project. The C-ORAL-BRASIL stemmed from the European C-ORAL-ROM project (Cresti & Moneglia, 2005), which has compiled spoken corpora of Italian, French, Spanish, and European Portuguese. The corpora of the C-ORAL family represent adequate tools for the analysis of spoken language, for they are provided not only with the transcripts of the recorded sessions (with prosodic breaks’ annotation), but also with their audio files and the text-to-speech alignment. So far, the C-ORAL-BRASIL project has published the C-ORAL-BRASIL I (Informal corpus: Raso & Mello, 2012), while the C-ORAL-BRASIL II (to be published by 2019) comprises a Formal corpus (Natural context), a Media corpus, and a Telephonic corpus. Besides these resources, a set of informationally tagged comparable minicorpora (representative samples of the aforementioned corpora) are already available or in preparation, enabling (cross-linguistic) studies focussed on information structure.

Download Full-text

Deep Learning Based TTS-STT Model with Transliteration for Indic Languages

International Journal for Research in Applied Science and Engineering Technology ◽

10.22214/ijraset.2021.39689 ◽

2021 ◽

Vol 9 (12) ◽

pp. 2207-2213

Author(s):

Kartik Tiwari

Keyword(s):

Open Source ◽

Speech Processing ◽

High Performance ◽

Translation System ◽

Data Sets ◽

Text To Speech ◽

Strong Focus ◽

End To End ◽

Translation Methods ◽

Speech Presentation

Abstract: This paper introduces a new text-to-speech presentation from end-to-end (E2E-TTS) using toolkit called ESPnet-TTS, which is an open source extension. ESPnet speech processing tools kit. Various models come under ESPnet TTS TacoTron 2, Transformer TTS, and Fast Speech. This also provides recipes recommended by the Kaldi speech recognition tool kit (ASR). Recipes based on the composition combined with the ESPnet ASR recipe, which provides high performance. This toolkit also provides pre-trained models and samples of all recipes for users to use as a base .It works on TTS-STT and translation features for various indicator languages, with a strong focus on English, Marathi and Hindi. This paper also shows that neural sequence-to-sequence models find the state of the art or near the effects of the art state on existing databases. We also analyze some of the key design challenges that contribute to the development of a multilingual business translation system, which includes processing bilingual business data sets and evaluating multiple translation methods. The test result can be obtained using tokens and these test results show that our models can achieve modern performance compared to the latest LJ Speech tool kit data. Terms of Reference — Open source, end-to-end, text-to-speech

Download Full-text

A Rule-Based Method for Homograph Disambiguation in Brazilian Portuguese Text-to-Speech Systems

Journal of Communication and Information Systems ◽

10.14209/jcis.2012.1 ◽

2012 ◽

Vol 27 (1) ◽

pp. 1-9 ◽

Cited By ~ 1

Author(s):

D.C. Silva ◽

D. Braga ◽

F.G.V. Resende Júnior

Keyword(s):

Portuguese Text ◽

Brazilian Portuguese ◽

Text To Speech ◽

Rule Based

Download Full-text