Entry Chen:2008:TTR from talip.bib
Last update: Sun Oct 15 02:55:04 MDT 2017
Top |
Symbols |
Numbers |
Math |
A |
B |
C |
D |
E |
F |
G |
H |
I |
J |
K |
L |
M |
N |
O |
P |
Q |
R |
S |
T |
U |
V |
W |
X |
Y |
Z
BibTeX entry
@Article{Chen:2008:TTR,
author = "Jiang-Chun Chen and Jyh-Shing Roger Jang",
title = "{TRUES}: {Tone Recognition Using Extended Segments}",
journal = j-TALIP,
volume = "7",
number = "3",
pages = "10:1--10:??",
month = aug,
year = "2008",
CODEN = "????",
DOI = "https://doi.org/10.1145/1386869.1386872",
ISSN = "1530-0226 (print), 1558-3430 (electronic)",
ISSN-L = "1530-0226",
bibdate = "Fri Aug 22 13:11:51 MDT 2008",
bibsource = "http://portal.acm.org/;
http://www.math.utah.edu/pub/tex/bib/talip.bib",
abstract = "Tone recognition has been a basic but important task
for speech recognition and assessment of tonal
languages, such as Mandarin Chinese. Most previously
proposed approaches adopt a two-step approach where
syllables within an utterance are identified via forced
alignment first, and tone recognition using a variety
of classifiers---such as neural networks, Gaussian
mixture models (GMM), hidden Markov models (HMM),
support vector machines (SVM)---is then performed on
each segmented syllable to predict its tone. However,
forced alignment does not always generate accurate
syllable boundaries, leading to unstable
voiced-unvoiced detection and deteriorating performance
in tone recognition. Aiming to alleviate this problem,
we propose a robust approach called Tone Recognition
Using Extended Segments (TRUES) for HMM-based
continuous tone recognition. The proposed approach
extracts an unbroken pitch contour from a given
utterance based on dynamic programming over time-domain
acoustic features of average magnitude difference
function (AMDF). The pitch contour of each syllable is
then extended for tri-tone HMM modeling, such that the
influence from inaccurate syllable boundaries is
lessened. Our experimental results demonstrate that the
proposed TRUES achieves 49.13\% relative error rate
reduction over that of the recently proposed supratone
modeling, which is deemed the state of the art of tone
recognition that outperforms several previously
proposed approaches. The encouraging improvement
demonstrates the effectiveness and robustness of the
proposed TRUES, as well as the corresponding pitch
determination algorithm which produces unbroken pitch
contours.",
acknowledgement = ack-nhfb,
articleno = "10",
fjournal = "ACM Transactions on Asian Language Information
Processing",
journal-URL = "http://portal.acm.org/browse_dl.cfm?&idx=J820",
keywords = "context-dependent tone modeling; continuous tone
recognition; extended segment for tone recognition;
HMM; Mandarin Chinese; supratone modeling",
}
Related entries
- accurate,
6(4)2,
12(3)12
- achieve,
7(3)9,
7(4)12,
7(4)13,
8(1)4,
8(2)7,
8(3)10,
8(4)17,
9(2)6,
9(4)14,
10(2)7,
11(3)8,
11(3)11,
11(4)15,
11(4)17,
12(1)3,
12(1)4,
12(2)7,
12(3)9,
12(3)11,
13(1)3,
13(1)4,
13(2)9
- acoustic,
9(2)7,
10(1)6,
10(2)7
- adopt,
11(2)7,
11(4)14,
13(3)13
- algorithm,
1(4)281,
5(2)165,
6(3)11,
6(4)2,
6(4)3,
7(1)3,
7(2)7,
7(3)8,
7(4)12,
8(1)4,
8(3)12,
9(1)3,
9(1)4,
10(2)7,
10(2)9,
10(2)10,
10(4)19,
11(1)3,
11(2)4,
11(2)7,
11(3)9,
11(3)11,
11(4)14,
12(1)1,
12(1)2,
12(2)6,
13(1)4,
13(3)13,
13(4)18
- alignment,
5(2)121,
7(1)1,
7(4)12,
8(2)6,
8(2)7,
8(2)8,
11(4)18,
12(3)9,
12(4)16
- alleviate,
6(3)9,
8(2)9,
9(1)1
- always,
9(3)11,
9(3)12,
10(1)6
- art,
12(3)9
- assessment,
9(3)12
- average,
9(3)11,
9(3)12,
11(3)10,
13(2)6
- based,
1(3)269,
2(4)301,
3(2)113,
4(3)357,
4(4)377,
5(2)121,
5(2)165,
5(3)185,
6(2)8,
6(3)9,
6(3)11,
6(4)3,
7(2)6,
7(4)12,
8(1)4,
8(3)11,
8(3)12,
8(4)17,
8(4)19,
9(1)2,
9(2)5,
9(2)6,
9(2)7,
9(3)11,
10(1)4,
10(1)6,
10(2)7,
10(3)14,
10(3)16,
11(2)6,
11(3)8,
11(3)11,
11(4)14,
11(4)15,
11(4)16,
11(4)17,
11(4)18,
12(1)1,
12(1)2,
12(1)4,
12(3)9,
12(3)10,
12(3)11,
13(1)2,
13(1)3,
13(1)4,
13(2)10,
13(3)13,
13(4)17,
13(4)18
- based, HMM-,
13(3)14
- basic,
8(1)4,
9(1)4,
10(3)15,
11(4)16,
13(3)14
- been,
5(2)121,
6(3)10,
6(4)3,
7(4)11,
7(4)12,
7(4)13,
8(1)2,
8(2)9,
8(4)14,
8(4)17,
8(4)18,
9(1)1,
9(2)5,
9(4)15,
10(1)5,
10(4)17,
13(1)1,
13(2)6,
13(2)8,
13(2)9,
13(3)11,
13(3)12
- boundary,
8(2)7,
10(1)6,
12(1)2
- called,
6(2)8,
7(3)9,
8(3)12,
12(1)1,
12(3)12,
13(1)2,
13(4)18
- classifier,
6(4)3,
8(4)15,
10(2)9,
10(4)17,
12(1)4,
12(2)7,
12(3)9,
13(3)12
- continuous,
1(1)83,
8(1)2
- contour,
3(3)169,
6(2)8
- corresponding,
5(2)89,
8(2)7,
9(2)7,
9(3)12,
10(2)7,
11(2)5,
11(3)8
- demonstrate,
5(2)121,
6(4)1,
7(1)2,
7(2)7,
8(2)7,
8(2)9,
9(1)4,
9(3)11,
11(2)6,
11(3)8,
13(1)1,
13(1)3
- detection,
2(2)85,
6(3)11,
8(4)16,
8(4)17,
9(1)2,
10(1)5,
10(1)6,
10(4)20,
10(4)21,
11(4)17,
13(2)10,
13(4)16
- determination,
1(3)269,
4(4)417,
11(2)5
- difference,
8(4)18,
9(1)3,
9(2)5,
9(3)11,
10(1)2,
10(4)17,
12(1)1,
12(3)11,
13(2)10,
13(4)16
- do,
7(3)8,
7(4)11,
7(4)12,
7(4)13,
8(2)7,
8(2)9,
12(1)3,
13(2)9
- dynamic,
10(3)15,
11(2)5
- each,
5(2)121,
5(2)146,
5(2)165,
7(3)8,
7(4)11,
7(4)13,
8(4)16,
8(4)18,
9(1)2,
10(1)2,
10(2)7,
10(2)9,
10(4)20,
12(1)2,
12(1)3,
12(3)9,
12(3)10,
13(1)2,
13(2)9,
13(3)13,
13(3)14,
13(4)17
- effectiveness,
6(2)6,
6(2)8,
6(4)2,
7(2)5,
8(2)7,
8(4)17,
9(2)5,
9(3)11,
9(3)12,
10(3)15,
11(2)5,
11(4)18,
12(4)16,
12(4)17,
13(2)6,
13(3)13,
13(4)18
- encouraging,
8(1)3,
11(2)5,
12(3)11
- error,
4(1)18,
6(3)9,
7(1)2,
9(1)2,
9(2)6,
10(1)2,
10(1)5,
10(1)6,
10(2)7,
10(2)10,
11(1)3,
11(2)7,
11(4)18,
12(1)2,
13(2)8,
13(3)14
- experimental,
5(2)121,
6(2)6,
6(2)7,
7(2)5,
7(2)6,
7(2)7,
7(4)13,
8(1)2,
8(1)4,
8(2)9,
9(1)1,
9(1)2,
10(2)7,
11(3)9,
11(4)14,
12(3)10,
12(4)14,
13(2)7
- extended,
5(3)183,
7(4)13,
8(4)15,
8(4)18,
9(1)3,
11(2)6,
13(4)16
- extract,
6(3)11,
7(3)8,
9(1)1,
9(2)6,
11(1)2,
11(2)4,
11(2)6,
11(3)11,
11(4)15,
12(2)7,
12(3)9
- feature,
2(3)290,
5(2)165,
6(4)1,
7(2)6,
7(2)7,
7(4)13,
8(3)11,
8(4)14,
8(4)17,
9(1)2,
9(2)5,
9(2)6,
10(1)5,
10(1)6,
10(2)7,
10(3)13,
10(3)15,
10(4)17,
10(4)19,
10(4)21,
11(3)10,
11(4)14,
11(4)16,
12(1)1,
12(1)4,
12(3)9,
12(3)10,
13(2)9,
13(3)13,
13(4)16
- first,
5(2)165,
6(2)6,
6(4)3,
7(1)1,
7(3)8,
8(2)7,
8(3)10,
8(3)11,
8(4)19,
9(3)10,
9(3)11,
10(3)13,
11(2)6,
11(2)7,
11(3)8,
11(3)9,
11(3)11,
12(1)3,
12(1)4,
12(2)5,
12(3)10,
12(4)17,
13(2)7,
13(2)9,
13(3)13,
13(4)17
- forced,
12(3)9
- function,
5(2)121,
7(3)8,
7(4)11,
11(1)3,
11(2)4
- generate,
3(3)190,
5(2)121,
8(1)4,
10(3)16,
12(3)9,
12(4)15,
13(4)16
- given,
5(2)89,
5(2)165,
7(4)13,
8(1)2,
9(1)1,
9(2)7,
11(2)5,
11(3)8,
11(3)11,
11(4)13,
12(4)14,
13(1)2,
13(4)17
- has,
5(2)89,
5(2)121,
6(4)2,
6(4)3,
7(4)11,
7(4)12,
7(4)13,
8(1)2,
8(3)10,
8(4)16,
8(4)17,
8(4)18,
9(1)1,
9(1)2,
9(1)3,
9(2)5,
9(4)15,
10(1)4,
10(1)5,
10(4)17,
11(1)2,
11(3)9,
12(1)1,
12(1)2,
12(4)15,
13(1)1,
13(2)6,
13(2)7,
13(2)8,
13(3)12,
13(4)18
- hidden,
3(2)146,
8(2)6,
8(3)12,
9(2)7,
10(4)20,
11(3)9,
13(3)12
- HMM,
4(1)38,
11(3)9,
13(3)14
- HMM-based,
13(3)14
- however,
6(2)8,
7(2)7,
7(3)8,
7(4)11,
7(4)12,
7(4)13,
8(3)10,
9(1)1,
9(3)12,
10(3)12,
10(4)18,
11(2)6,
11(2)7,
12(1)2,
12(2)7,
12(3)10,
12(3)11,
12(4)14,
13(3)11,
13(4)17
- identified,
8(3)11,
13(3)12
- important,
5(4)413,
7(1)3,
9(1)1,
9(2)7,
9(4)13,
10(2)10,
11(4)18,
12(1)3,
13(3)12,
13(4)18
- improvement,
4(3)280,
5(2)121,
6(2)7,
7(1)2,
7(1)3,
7(3)8,
8(1)4,
8(3)10,
8(4)15,
8(4)16,
9(1)3,
9(3)11,
10(2)8,
11(4)17,
11(4)18,
12(1)1,
12(2)7,
12(3)11,
13(3)12,
13(3)14,
13(4)16,
13(4)17
- influence,
8(2)7,
12(3)11
- leading,
10(2)9,
10(3)15
- machine,
4(1)18,
4(4)377,
5(2)89,
5(3)185,
6(2)6,
6(4)2,
6(4)3,
7(1)1,
7(2)5,
7(2)7,
7(3)9,
8(1)4,
8(2)5,
8(2)6,
8(2)7,
8(2)8,
8(2)9,
8(3)10,
8(4)15,
9(1)3,
9(4)13,
10(1)2,
10(1)5,
10(3)16,
10(4)18,
10(4)20,
11(1)2,
11(3)8,
11(4)14,
11(4)16,
12(3)9,
12(3)12,
12(4)14,
12(4)16,
12(4)17,
13(1)2,
13(3)11,
13(4)16,
13(4)17
- Mandarin,
3(2)128,
5(2)89,
8(1)2,
8(1)3,
9(1)2
- Markov,
3(2)146,
6(2)7,
8(2)6,
11(3)9,
11(4)15,
13(3)12
- mixture,
5(2)165
- modeling,
1(1)3,
1(3)173,
3(2)87,
3(3)169,
6(1)z,
6(2)6,
6(3)9,
7(4)13,
8(1)2,
8(1)4,
9(4)14,
10(4)18,
10(4)21,
11(2)5,
12(2)5,
13(3)12,
13(4)16
- most,
6(2)6,
7(1)1,
7(3)8,
8(4)15,
9(1)1,
9(2)5,
9(3)11,
10(1)5,
10(2)8,
12(1)1,
12(1)2,
13(1)1,
13(1)4,
13(2)6,
13(3)14,
13(4)18
- network,
4(1)38,
8(1)4,
11(4)15,
13(4)16,
13(4)18
- neural,
4(1)38
- not,
5(2)89,
5(2)121,
6(3)10,
7(2)7,
7(3)8,
7(4)11,
7(4)12,
7(4)13,
8(1)3,
8(2)7,
8(2)8,
8(2)9,
8(4)17,
9(2)6,
9(2)7,
9(4)14,
10(1)4,
10(1)5,
10(2)8,
10(2)10,
10(4)18,
11(1)2,
11(3)10,
12(1)2,
12(1)3,
12(2)7,
12(3)9,
13(2)7,
13(2)9,
13(4)17,
13(4)18
- outperform,
6(4)3,
7(2)6,
7(4)13,
8(2)6,
9(2)5,
9(2)6,
10(3)15,
11(2)7,
11(3)8,
11(4)14,
12(2)5,
13(3)13,
13(4)17
- performance,
5(2)121,
5(2)165,
6(2)8,
6(3)9,
6(4)1,
6(4)3,
7(1)1,
7(1)2,
7(2)5,
7(2)6,
7(2)7,
7(3)9,
7(4)13,
8(1)2,
8(1)3,
8(2)7,
8(2)8,
8(2)9,
8(3)10,
8(4)16,
8(4)17,
8(4)18,
9(1)2,
9(1)4,
9(2)5,
9(2)6,
9(3)11,
9(3)12,
9(4)14,
10(2)8,
10(3)13,
10(3)14,
11(2)7,
11(3)10,
11(3)11,
11(4)14,
11(4)15,
11(4)17,
12(1)2,
12(3)9,
12(3)11,
12(4)14,
12(4)15,
12(4)16,
13(1)3,
13(1)4,
13(2)7,
13(2)9,
13(4)16,
13(4)17
- performed,
8(1)2,
9(3)11,
10(1)2,
12(2)7,
13(2)6,
13(4)17
- predict,
12(4)15,
13(2)8,
13(2)9
- previously,
12(4)17
- problem,
6(2)7,
6(3)9,
6(3)11,
6(4)1,
7(1)2,
7(2)7,
8(1)2,
8(2)9,
8(3)10,
8(4)19,
9(1)1,
9(1)3,
9(2)5,
9(4)13,
10(1)2,
10(1)4,
10(3)14,
10(3)16,
10(4)21,
11(3)8,
11(3)11,
11(4)17,
11(4)18,
12(1)2,
12(1)3,
12(2)7,
12(3)10,
12(3)12,
12(4)16,
13(2)8,
13(3)14,
13(4)17
- produce,
10(2)8,
11(1)3
- programming,
11(1)3,
12(1)3
- propose,
5(2)89,
6(2)6,
6(2)8,
6(3)11,
7(3)8,
7(4)12,
8(1)2,
8(1)4,
8(2)9,
8(4)19,
9(2)7,
9(4)13,
10(2)10,
10(3)12,
10(3)15,
10(4)17,
10(4)20,
11(2)7,
11(3)9,
11(3)11,
11(4)15,
11(4)16,
11(4)18,
12(1)1,
12(1)2,
12(1)3,
12(1)4,
12(2)5,
12(2)6,
12(3)9,
12(3)10,
12(3)12,
12(4)16,
13(1)2,
13(1)3,
13(2)8,
13(2)9,
13(3)12,
13(3)13,
13(4)17,
13(4)18
- proposed,
5(2)121,
5(2)165,
6(2)7,
7(1)1,
7(1)2,
7(3)9,
7(4)11,
7(4)13,
8(1)4,
8(2)6,
8(3)10,
8(3)11,
8(4)14,
8(4)19,
9(1)1,
9(2)5,
9(2)7,
10(2)7,
10(2)9,
10(3)14,
10(4)18,
11(1)3,
11(2)5,
11(2)6,
11(2)7,
11(3)8,
11(3)9,
11(3)10,
11(3)11,
11(4)16,
11(4)17,
12(1)4,
12(2)5,
12(2)7,
12(3)12,
12(4)16,
12(4)17,
13(2)6,
13(2)8,
13(3)13,
13(4)18
- rate,
6(3)9,
7(1)2,
10(1)2,
10(1)6,
10(2)7,
12(1)1,
12(1)4,
13(2)8
- recently,
11(1)2,
13(1)1
- recognition,
1(1)83,
1(4)297,
2(1)27,
2(3)290,
5(1)4,
5(2)165,
6(2)6,
6(3)9,
6(4)3,
7(1)2,
8(1)2,
8(3)11,
8(4)18,
9(1)2,
9(2)7,
10(1)6,
10(2)7,
10(2)9,
10(3)13,
11(1)1,
11(1)2,
11(4)13,
11(4)16,
11(4)17,
11(4)18,
12(1)4,
12(3)10,
13(3)12,
13(4)16
- reduction,
3(2)146,
7(1)2,
7(1)3,
10(2)7,
11(3)10,
13(1)3,
13(3)12
- relative,
8(2)7,
9(3)11,
12(1)2,
12(3)9,
13(1)2
- result,
4(2)135,
5(2)121,
5(2)146,
5(2)165,
6(2)6,
6(2)7,
6(3)9,
6(3)11,
6(4)3,
7(1)2,
7(2)5,
7(2)6,
7(2)7,
7(3)8,
7(4)11,
7(4)12,
7(4)13,
8(1)2,
8(1)3,
8(1)4,
8(2)6,
8(2)9,
8(3)10,
8(3)12,
8(4)14,
8(4)15,
8(4)16,
8(4)17,
8(4)18,
8(4)19,
9(1)1,
9(1)2,
9(2)5,
9(2)6,
9(2)7,
9(3)11,
9(3)12,
9(4)14,
10(1)2,
10(2)7,
11(2)4,
11(2)5,
11(3)8,
11(3)9,
11(3)11,
11(4)13,
11(4)14,
11(4)15,
12(1)3,
12(1)4,
12(2)5,
12(2)7,
12(3)9,
12(3)10,
12(3)11,
12(4)14,
12(4)16,
13(1)1,
13(1)4,
13(2)6,
13(2)7,
13(2)9,
13(3)11,
13(3)12,
13(3)14
- robust,
5(2)89,
7(1)2,
11(2)7,
12(1)4,
12(2)5,
12(2)7,
13(3)12,
13(4)18
- robustness,
7(1)2,
11(2)7
- segment,
9(1)2,
12(1)2,
12(1)4
- segmented,
8(2)7,
8(3)11,
9(2)5,
12(1)4,
13(2)9
- several,
6(2)6,
6(2)7,
6(4)3,
7(2)5,
7(2)7,
8(3)10,
8(4)16,
8(4)17,
8(4)18,
9(3)12,
11(2)6,
11(4)13,
11(4)16,
12(1)2,
13(3)12,
13(3)14
- speech,
1(1)83,
4(1)38,
6(3)9,
7(1)2,
8(1)2,
8(1)4,
8(4)14,
8(4)18,
9(1)2,
9(2)7,
10(1)6,
10(2)7,
11(1)2,
11(3)10,
13(4)16
- state,
5(2)165,
9(3)8,
11(3)9,
12(3)9,
13(3)14,
13(3)15
- step, two-,
6(2)6,
12(3)12
- such,
7(2)7,
7(3)8,
7(4)12,
8(2)8,
8(3)10,
8(3)11,
8(3)12,
8(4)14,
8(4)16,
8(4)17,
9(1)1,
9(3)12,
9(4)13,
9(4)15,
10(1)5,
10(2)8,
10(3)12,
10(4)21,
11(1)2,
11(2)5,
11(2)7,
11(3)8,
11(3)10,
11(3)11,
11(4)13,
11(4)16,
11(4)17,
11(4)18,
12(1)1,
12(1)2,
12(2)6,
12(3)10,
12(3)11,
12(4)14,
12(4)17,
13(1)1,
13(3)12,
13(4)17
- support,
6(4)3,
7(2)7,
7(3)8,
12(2)6,
12(3)9,
13(1)4,
13(4)18
- SVM,
7(2)7,
8(3)11,
12(1)4
- syllable,
10(1)6,
10(2)7,
12(1)2
- task,
2(1)49,
5(2)89,
5(2)121,
6(2)7,
6(3)11,
6(4)1,
6(4)3,
7(1)1,
7(1)2,
7(2)7,
7(4)13,
8(1)4,
8(2)7,
8(4)15,
8(4)16,
9(1)4,
9(2)6,
9(3)10,
9(4)14,
9(4)15,
10(1)5,
10(3)14,
10(4)18,
10(4)20,
10(4)21,
11(1)2,
11(3)8,
11(3)11,
11(4)13,
11(4)14,
11(4)17,
11(4)18,
12(1)2,
12(1)3,
12(2)5,
12(2)7,
12(3)9,
12(4)17,
13(2)10,
13(4)17
- then,
5(2)121,
6(2)6,
7(1)1,
7(4)12,
8(1)4,
8(2)7,
8(3)10,
8(3)11,
8(3)12,
8(4)14,
9(1)1,
9(2)7,
9(3)11,
10(2)7,
10(3)13,
10(4)20,
11(1)3,
11(2)7,
11(3)11,
11(4)15,
12(1)3,
12(3)10,
12(4)17,
13(1)4,
13(2)9,
13(3)13,
13(4)16
- tone,
1(1)83,
1(3)207
- two-step,
6(2)6,
12(3)12
- unstable,
13(4)17
- variety,
8(2)6,
8(4)18,
11(1)2
- vector,
6(4)3,
7(2)7,
10(4)19,
11(2)5,
12(3)9
- via,
1(1)34,
4(2)111,
5(2)146,
7(1)1,
8(2)6,
10(3)12,
10(3)15,
11(2)5,
12(4)15,
12(4)17,
13(3)12
- well,
5(2)121,
6(2)6,
6(3)11,
7(3)8,
7(3)9,
8(1)2,
8(4)18,
9(1)2,
9(3)12,
10(3)15,
11(1)2,
11(2)4,
11(3)11,
12(3)9,
13(1)2
- where,
7(3)8,
7(3)9,
8(2)7,
8(2)8,
9(1)3,
9(2)7,
9(4)13,
10(4)18,
11(2)4,
11(4)16,
12(1)3,
12(4)14
- which,
5(2)89,
5(2)121,
6(2)6,
6(2)8,
6(3)10,
6(4)1,
7(1)3,
7(2)7,
7(4)13,
8(1)2,
8(1)4,
8(2)9,
8(3)10,
8(3)12,
8(4)14,
8(4)16,
8(4)18,
9(1)1,
9(2)5,
9(3)12,
9(4)14,
9(4)15,
10(1)4,
10(1)6,
10(2)7,
10(2)8,
10(2)9,
10(3)12,
10(3)13,
10(3)14,
10(3)15,
10(4)17,
10(4)19,
10(4)20,
11(2)6,
11(3)8,
11(3)9,
11(4)13,
11(4)14,
11(4)15,
11(4)16,
11(4)18,
12(1)1,
12(1)3,
12(1)4,
12(2)5,
12(2)6,
12(2)7,
12(3)10,
12(3)11,
12(4)15,
12(4)17,
13(1)1,
13(2)8,
13(3)13,
13(4)17,
13(4)18