Sign in
Author
|
Conference
|
Journal
|
Organization
|
Year
|
DOI
Look for results that meet for the following criteria:
since
equal to
before
between
and
Search in all fields of study
Limit my searches in the following fields of study
Agriculture Science
Arts & Humanities
Biology
Chemistry
Computer Science
Economics & Business
Engineering
Environmental Sciences
Geosciences
Material Science
Mathematics
Medicine
Physics
Social Science
Multidisciplinary
Keywords
(14)
bayesian analysis
bayesian method
Design Methodology
Discrimination Learning
Discriminative Training
Error Analysis
Functional Form
Integrated Design
Language Model
Machine Translation
Speech Recognition
Speech Translation
Decision Feedback
Minimum Classification Error
Subscribe
Academic
Publications
A novel decision function and the associated decision-feedback learning for speech translation
A novel decision function and the associated decision-feedback learning for speech translation,10.1109/ICASSP.2011.5947631,Yaodong Zhang,Li Deng,Xiaod
Edit
A novel decision function and the associated decision-feedback learning for speech translation
(
Citations: 5
)
BibTex
|
RIS
|
RefWorks
Download
Yaodong Zhang
,
Li Deng
,
Xiaodong He
,
Alex Acero
In this paper we report our recent development of an end-to-end integrative
design methodology
for speech translation. Specifically, a novel decision function is proposed based on the Bayesian analysis, and the associated discriminative learning technique is presented based on the decision-feedback principle. The decision function in our end-to-end
design methodology
integrates acoustic scores,
language model
scores and translation scores to refine the translation hypotheses and to determine the best translation candidate. This Bayesian-guided decision function is then embedded into the training process that jointly learns the parameters in
speech recognition
and
machine translation
sub-systems in the overall
speech translation
system. The resulting decision-feedback learning takes a
functional form
similar to the
minimum classification error
training. Experimental results obtained on the IWSLT DIALOG 2010 database showed that the proposed system outperformed the baseline system in terms of BLEU score by 2.3 points.
Conference:
International Conference on Acoustics, Speech, and Signal Processing - ICASSP
, pp. 5608-5611, 2011
DOI:
10.1109/ICASSP.2011.5947631
Cumulative
Annual
View Publication
The following links allow you to view full publications. These links are maintained by other sources not affiliated with Microsoft Academic Search.
(
research.microsoft.com
)
(
research.microsoft.com
)
(
ieeexplore.ieee.org
)
(
ieeexplore.ieee.org
)
More »
Citation Context
(4)
...We have demonstrated that such an end-to-end optimization approach can give better results for complex information systems such speech translation involoving multiple sub-systems working together [22,
23
]...
Gokhan Tur
,
et al.
Towards Deeper Understanding Deep Convex Networks for Semantic Utteran...
...The complexity of the end-to-end optimization task requires careful optimization techniques [3,4,19,
20
].,It is desirable to optimize these downstream components in a consistent way and with the same end-toend performance measure, which we address in this section [
20
]...
Xiaodong He
,
et al.
Optimization in Speech-Centric Information Processing: Criteria and te...
...In the parallel work reported in [
18
], we adopt a more aggressive approach where the parameters “inside” the feature functions, e.g., all the individual N-gram probability values of both source and target languages, as well as the phrase table probability values, are subject to adjustment so as to maximize the end-to-end ST quality...
Xiaodong He
,
et al.
Why word error rate is not a good metric for speech recognizer trainin...
...Our recent work on speech translation (e.g., [1, 16,
17
, 18]) pioneered the direction of end-to-end discriminative learning for the entire speech translation system design...
Xiaodong He
,
et al.
Robust Speech Translation by Domain Adaptation
References
(13)
Improved speech recognition word lattice translation by confidence measure
(
Citations: 3
)
Abdulvohid Bozarov
,
Yoshinori Sagisaka
,
Ruiqiang Zhang
,
Gen-ichiro Kikui
Published in 2005.
Why word error rate is not a good metric for speech recognizer training for the speech translation task?
(
Citations: 4
)
Xiaodong He
,
Li Deng
,
Alex Acero
Conference:
International Conference on Acoustics, Speech, and Signal Processing - ICASSP
, pp. 5632-5635, 2011
Minimum classification error rate methods for speech recognition
(
Citations: 319
)
Biing-Hwang Juang
,
Wu Hou
,
Chin-Hui Lee
Journal:
IEEE Transactions on Speech and Audio Processing - IEEE SAP
, vol. 5, no. 3, pp. 257-265, 1997
Statistical Phrase-Based Translation
(
Citations: 862
)
Philipp Koehn
,
Franz Josef Och
,
Daniel Marcu
Conference:
North American Chapter of the Association for Computational Linguistics - NAACL
, 2003
Integrating Speech Recognition and Machine Translation: Where do We Stand?
(
Citations: 5
)
Evgeny Matusov
,
Stephan Kanthak
,
Hermann Ney
Conference:
International Conference on Acoustics, Speech, and Signal Processing - ICASSP
, vol. 5, pp. V-V, 2006
Sort by:
Citations
(5)
Towards Deeper Understanding Deep Convex Networks for Semantic Utterance Classification
(
Citations: 3
)
Gokhan Tur
,
Li Deng
,
Dilek Hakkani-Tur
,
Xiaodong He
Published in 2012.
Optimization in Speech-Centric Information Processing: Criteria and techniques
Xiaodong He
,
Li Deng
Published in 2012.
Why word error rate is not a good metric for speech recognizer training for the speech translation task?
(
Citations: 4
)
Xiaodong He
,
Li Deng
,
Alex Acero
Conference:
International Conference on Acoustics, Speech, and Signal Processing - ICASSP
, pp. 5632-5635, 2011
Robust Speech Translation by Domain Adaptation
(
Citations: 1
)
Xiaodong He
,
Li Deng
Published in 2011.
s p eech r e cognition, Machine t r anslation, and s p eech t r anslation—A u n ified di scriminative Learning Paradigm
Unknown
Published in 2011.