Show simple item record

Depositordc.contributorEspic, Felipe
Funderdc.contributor.otherEPSRC - Engineering and Physical Sciences Research Councilen_UK
Data Creatordc.creatorEspic, Felipe
Data Creatordc.creatorValentini-Botinhao, Cassia
Data Creatordc.creatorWu, Zhizheng
Data Creatordc.creatorKing, Simon
Date Accessioneddc.date.accessioned2016-06-24T09:37:44Z
Date Availabledc.date.available2016-06-24T09:37:44Z
Citationdc.identifier.citationEspic, Felipe; Valentini-Botinhao, Cassia; Wu, Zhizheng; King, Simon. (2016). Listening test materials for "Waveform generation based on signal reshaping for statistical parametric speech synthesis", [dataset]. University of Edinburgh. The Centre for Speech Technology Research (CSTR). https://doi.org/10.7488/ds/1433.en
Persistent Identifierdc.identifier.urihttp://hdl.handle.net/10283/2046
Persistent Identifierdc.identifier.urihttps://doi.org/10.7488/ds/1433
Dataset Description (abstract)dc.description.abstractThe dataset contains the testing stimuli and listeners' MUSHRA test responses for the Interspeech 2016 paper "Waveform generation based on signal reshaping for statistical parametric speech synthesis". On this paper, we propose a new paradigm of waveform generation for Statistical Parametric Speech Synthesis that is based on neither source-filter separation nor sinusoidal modelling. We suggest that one of the main problems of current vocoding techniques is that they perform an extreme decomposition of the speech signal into source and filter, which is an underlying cause of "buzziness", "musical artifacts", or "muffled sound" in the synthetic speech. The proposed method avoids making unnecessary assumptions and decompositions as far as possible, and uses only the spectral envelope and F0 as parameters. Pre-recorded speech is used as a base signal, which is "reshaped" to match the acoustic specification predicted by the statistical model, without any source-filter decomposition. A detailed description of the method is presented, including implementation details and adjustments. Subjective listening test evaluations of complete DNN-based text-to-speech systems were conducted for two voices: one female and one male. The results show that the proposed method tends to outperform the state-of-the-art standard vocoder STRAIGHT, whilst using fewer acoustic parameters.en_UK
Languagedc.language.isoengen_UK
Publisherdc.publisherUniversity of Edinburgh. The Centre for Speech Technology Research (CSTR)en_UK
Rightsdc.rightsCreative Commons Attribution 4.0 International Public Licenseen
Subjectdc.subjectwaveform generationen_UK
Subjectdc.subjectspeech synthesisen_UK
Subjectdc.subjectvocoderen_UK
Subject Classificationdc.subject.classificationMathematical and Computer Sciences::Speech and Natural Language Processingen_UK
Titledc.titleListening test materials for "Waveform generation based on signal reshaping for statistical parametric speech synthesis"en_UK
Typedc.typedataseten_UK

Download All
zip file MD5 Checksum: b4719bc0c687ec9baba6d97c272960aa

Files in this item

Thumbnail
Thumbnail
Thumbnail
Thumbnail

This item appears in the following Collection(s)

Show simple item record