2026-09-18T21:22:18Zhttps://dspace.mit.edu/server/oai/requestoai:dspace.mit.edu:1721.1/169442022-01-13T07:54:29Zcom_1721.1_7582com_1721.1_7581col_1721.1_131022
James R. Glass.
Yi, Jon Rong-Wei, 1975-
Massachusetts Institute of Technology. Dept. of Electrical Engineering and Computer Science.
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
2005-05-19T15:22:53Z
2005-05-19T15:22:53Z
2003
2003
http://hdl.handle.net/1721.1/16944
53246622
Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2003.
Includes bibliographical references (p. 179-196).
This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.
Speech synthesis is an automatic encoding process carried out by machine through which symbols conveying linguistic information are converted into an acoustic waveform. In the past decade or so, a recent trend toward a non-parametric, corpus-based approach has focused on using real human speech as source material for producing novel natural-sounding speech. This work proposes a communication-theoretic formulation in which unit selection is a noisy channel through which an input sequence of symbols passes and an output sequence, possibly corrupted due to the coverage limits of the corpus, emerges. The penalty of approximation is quantified by substitution and concatenation costs which grade what unit contexts are interchangeable and where concatenations are not perceivable. These costs are semi-automatically derived from data and are found to agree with acoustic-phonetic knowledge. The implementation is based on a finite-state transducer (FST) representation that has been successfully used in speech and language processing applications including speech recognition. A proposed constraint kernel topology connects all units in the corpus with associated substitution and concatenation costs and enables an efficient Viterbi search that operates with low latency and scales to large corpora. An A* search can be applied in a second, rescoring pass to incorporate finer acoustic modelling. Extensions to this FST-based search include hierarchical and paralinguistic modelling. The search can also be used in an iterative feedback loop to record new utterances to enhance corpus coverage. This speech synthesis framework has been deployed across various domains and languages in many voices, a testament to its flexibility and rapid prototyping capability.
(cont.) Experimental subjects completing tasks in a given air travel planning scenario by interacting in real time with a spoken dialogue system over the telephone have found the system "easiest to understand" out of eight competing systems. In more detailed listening evaluations, subjective opinions garnered from human participants are found to be correlated with objective measures calculable by machine.
by Jon Rong-Wei Yi.
Ph.D.
214 p.
3462128 bytes
3461885 bytes
application/pdf
application/pdf
eng
Massachusetts Institute of Technology
M.I.T. theses are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission. See provided URL for inquiries about permission.
http://dspace.mit.edu/handle/1721.1/7582
Electrical Engineering and Computer Science.
Corpus-based unit selection for natural-sounding speech synthesis
Thesis
application/pdf
Publication
unknown
unknown
<Publication xmlns="https://www.openaire.eu/cerif-profile/1.1/" id="6c7e2aff-d2c6-442f-8a9b-a9dafca54c23">
<Type xmlns="https://www.openaire.eu/cerif-profile/vocab/COAR_Publication_Types">http://purl.org/coar/resource_type/c_1843</Type>
<Language>eng</Language>
<Title>Corpus-based unit selection for natural-sounding speech synthesis</Title>
<PublishedIn>
<Publication>
</Publication>
</PublishedIn>
<PublicationDate>2003</PublicationDate>
<Authors>
<Author>
<DisplayName>Yi, Jon Rong-Wei, 1975-</DisplayName>
<Affiliation>
<OrgUnit>
</OrgUnit>
</Affiliation>
</Author>
</Authors>
<Editors>
</Editors>
<Publishers>
<Publisher>
<DisplayName>Massachusetts Institute of Technology</DisplayName>
<OrgUnit />
</Publisher>
</Publishers>
<License>http://dspace.mit.edu/handle/1721.1/7582</License>
<Keyword>Electrical Engineering and Computer Science.</Keyword>
<Abstract>Speech synthesis is an automatic encoding process carried out by machine through which symbols conveying linguistic information are converted into an acoustic waveform. In the past decade or so, a recent trend toward a non-parametric, corpus-based approach has focused on using real human speech as source material for producing novel natural-sounding speech. This work proposes a communication-theoretic formulation in which unit selection is a noisy channel through which an input sequence of symbols passes and an output sequence, possibly corrupted due to the coverage limits of the corpus, emerges. The penalty of approximation is quantified by substitution and concatenation costs which grade what unit contexts are interchangeable and where concatenations are not perceivable. These costs are semi-automatically derived from data and are found to agree with acoustic-phonetic knowledge. The implementation is based on a finite-state transducer (FST) representation that has been successfully used in speech and language processing applications including speech recognition. A proposed constraint kernel topology connects all units in the corpus with associated substitution and concatenation costs and enables an efficient Viterbi search that operates with low latency and scales to large corpora. An A* search can be applied in a second, rescoring pass to incorporate finer acoustic modelling. Extensions to this FST-based search include hierarchical and paralinguistic modelling. The search can also be used in an iterative feedback loop to record new utterances to enhance corpus coverage. This speech synthesis framework has been deployed across various domains and languages in many voices, a testament to its flexibility and rapid prototyping capability.</Abstract>
<Abstract>(cont.) Experimental subjects completing tasks in a given air travel planning scenario by interacting in real time with a spoken dialogue system over the telephone have found the system "easiest to understand" out of eight competing systems. In more detailed listening evaluations, subjective opinions garnered from human participants are found to be correlated with objective measures calculable by machine.</Abstract>
<Access xmlns="http://purl.org/coar/access_right"
>
</Access>
</Publication>