Low latency parameter generation for real-time speech synthesis system

Xingyu Na; Xiang Xie; Jingming Kuang

doi:10.1109/ICME.2014.6890197

Low latency parameter generation for real-time speech synthesis system

Xingyu Na
, Xiang Xie
, Jingming Kuang

School of Information and Electronics

Beijing Institute of Technology

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

4 Citations (Scopus)

Abstract

Speech synthesizer is commonly used in human-computer interaction. In many applicational cases, the computing resource is limited while real-time synthesis is demanded. The HMM-based speech synthesis technique allows creating a natural voice quality with small footprint, but current synthesizers require the concatenation of sentence level acoustic units, which is not applicable in real-time mode. In this paper, we propose a blocked parameter generation algorithm for low latency speech synthesis which can work real-time in resource limited applications. Phonetic units at various time spans are used as blocks. The objective and subjective evaluations suggest that the proposed system produce promising voice quality with a low demand for the computing resource.

Original language	English
Title of host publication	2014 IEEE International Conference on Multimedia and Expo, ICME 2014
Publisher	IEEE Computer Society
Edition	Septmber
ISBN (Electronic)	9781479947614
DOIs	https://doi.org/10.1109/ICME.2014.6890197
Publication status	Published - 3 Sept 2014
Event	2014 IEEE International Conference on Multimedia and Expo, ICME 2014 - Chengdu, China Duration: 14 Jul 2014 → 18 Jul 2014

Publication series

Name	Proceedings - IEEE International Conference on Multimedia and Expo
Number	Septmber
Volume	2014-September
ISSN (Print)	1945-7871
ISSN (Electronic)	1945-788X

Conference

Conference	2014 IEEE International Conference on Multimedia and Expo, ICME 2014
Country/Territory	China
City	Chengdu
Period	14/07/14 → 18/07/14

Keywords

human computer interaction
real-time text-to-speech
speech synthesis

Access to Document

10.1109/ICME.2014.6890197

Cite this

Na, X., Xie, X., & Kuang, J. (2014). Low latency parameter generation for real-time speech synthesis system. In 2014 IEEE International Conference on Multimedia and Expo, ICME 2014 (Septmber ed.). Article 6890197 (Proceedings - IEEE International Conference on Multimedia and Expo; Vol. 2014-September, No. Septmber). IEEE Computer Society. https://doi.org/10.1109/ICME.2014.6890197

@inproceedings{1ffddf9180084850a125b9eb608ff6a6,

title = "Low latency parameter generation for real-time speech synthesis system",

abstract = "Speech synthesizer is commonly used in human-computer interaction. In many applicational cases, the computing resource is limited while real-time synthesis is demanded. The HMM-based speech synthesis technique allows creating a natural voice quality with small footprint, but current synthesizers require the concatenation of sentence level acoustic units, which is not applicable in real-time mode. In this paper, we propose a blocked parameter generation algorithm for low latency speech synthesis which can work real-time in resource limited applications. Phonetic units at various time spans are used as blocks. The objective and subjective evaluations suggest that the proposed system produce promising voice quality with a low demand for the computing resource.",

keywords = "human computer interaction, real-time text-to-speech, speech synthesis",

author = "Xingyu Na and Xiang Xie and Jingming Kuang",

note = "Publisher Copyright: {\textcopyright} 2014 IEEE.; 2014 IEEE International Conference on Multimedia and Expo, ICME 2014 ; Conference date: 14-07-2014 Through 18-07-2014",

year = "2014",

month = sep,

day = "3",

doi = "10.1109/ICME.2014.6890197",

language = "English",

series = "Proceedings - IEEE International Conference on Multimedia and Expo",

publisher = "IEEE Computer Society",

number = "Septmber",

booktitle = "2014 IEEE International Conference on Multimedia and Expo, ICME 2014",

address = "United States",

edition = "Septmber",

}

Na, X, Xie, X & Kuang, J 2014, Low latency parameter generation for real-time speech synthesis system. in 2014 IEEE International Conference on Multimedia and Expo, ICME 2014. Septmber edn, 6890197, Proceedings - IEEE International Conference on Multimedia and Expo, no. Septmber, vol. 2014-September, IEEE Computer Society, 2014 IEEE International Conference on Multimedia and Expo, ICME 2014, Chengdu, China, 14/07/14. https://doi.org/10.1109/ICME.2014.6890197

Low latency parameter generation for real-time speech synthesis system. / Na, Xingyu; Xie, Xiang; Kuang, Jingming.
2014 IEEE International Conference on Multimedia and Expo, ICME 2014. Septmber. ed. IEEE Computer Society, 2014. 6890197 (Proceedings - IEEE International Conference on Multimedia and Expo; Vol. 2014-September, No. Septmber).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Low latency parameter generation for real-time speech synthesis system

AU - Na, Xingyu

AU - Xie, Xiang

AU - Kuang, Jingming

PY - 2014/9/3

Y1 - 2014/9/3

N2 - Speech synthesizer is commonly used in human-computer interaction. In many applicational cases, the computing resource is limited while real-time synthesis is demanded. The HMM-based speech synthesis technique allows creating a natural voice quality with small footprint, but current synthesizers require the concatenation of sentence level acoustic units, which is not applicable in real-time mode. In this paper, we propose a blocked parameter generation algorithm for low latency speech synthesis which can work real-time in resource limited applications. Phonetic units at various time spans are used as blocks. The objective and subjective evaluations suggest that the proposed system produce promising voice quality with a low demand for the computing resource.

AB - Speech synthesizer is commonly used in human-computer interaction. In many applicational cases, the computing resource is limited while real-time synthesis is demanded. The HMM-based speech synthesis technique allows creating a natural voice quality with small footprint, but current synthesizers require the concatenation of sentence level acoustic units, which is not applicable in real-time mode. In this paper, we propose a blocked parameter generation algorithm for low latency speech synthesis which can work real-time in resource limited applications. Phonetic units at various time spans are used as blocks. The objective and subjective evaluations suggest that the proposed system produce promising voice quality with a low demand for the computing resource.

KW - human computer interaction

KW - real-time text-to-speech

KW - speech synthesis

UR - https://www.scopus.com/pages/publications/84937501080

U2 - 10.1109/ICME.2014.6890197

DO - 10.1109/ICME.2014.6890197

M3 - Conference contribution

AN - SCOPUS:84937501080

T3 - Proceedings - IEEE International Conference on Multimedia and Expo

BT - 2014 IEEE International Conference on Multimedia and Expo, ICME 2014

PB - IEEE Computer Society

T2 - 2014 IEEE International Conference on Multimedia and Expo, ICME 2014

Y2 - 14 July 2014 through 18 July 2014

ER -

Low latency parameter generation for real-time speech synthesis system

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this