Abstract
The application of Bayesian fusion of confidence measures to speech recognition is proposed. Feature level, decision level, and hybrid fusion are considered under the Bayesian framework. The use of speaker-adapted feature-level Bayesian fusion reduced the error rate by 19.4% as compared to the conventional single feature-based confidence scoring in an isolated word out-of-vocabulary rejection test. The decision-level Bayesian fusion also showed better performance than the majority rule. Finally, hybrid Bayesian fusion, which can combine both confidence measure features and local decisions, achieved the best performance.
| Original language | English |
|---|---|
| Pages (from-to) | 871-874 |
| Number of pages | 4 |
| Journal | IEEE Signal Processing Letters |
| Volume | 12 |
| Issue number | 12 |
| DOIs | |
| Publication status | Published - 2005 Dec 1 |
Fingerprint
Keywords
- Adaptive confidence scoring
- Bayesian fusion
- Confidence measure (CM)
- Speech recognition
ASJC Scopus subject areas
- Electrical and Electronic Engineering
- Signal Processing
Cite this
Bayesian fusion of confidence measures for speech recognition. / Kim, Tae Yoon; Ko, Hanseok.
In: IEEE Signal Processing Letters, Vol. 12, No. 12, 01.12.2005, p. 871-874.Research output: Contribution to journal › Article
}
TY - JOUR
T1 - Bayesian fusion of confidence measures for speech recognition
AU - Kim, Tae Yoon
AU - Ko, Hanseok
PY - 2005/12/1
Y1 - 2005/12/1
N2 - The application of Bayesian fusion of confidence measures to speech recognition is proposed. Feature level, decision level, and hybrid fusion are considered under the Bayesian framework. The use of speaker-adapted feature-level Bayesian fusion reduced the error rate by 19.4% as compared to the conventional single feature-based confidence scoring in an isolated word out-of-vocabulary rejection test. The decision-level Bayesian fusion also showed better performance than the majority rule. Finally, hybrid Bayesian fusion, which can combine both confidence measure features and local decisions, achieved the best performance.
AB - The application of Bayesian fusion of confidence measures to speech recognition is proposed. Feature level, decision level, and hybrid fusion are considered under the Bayesian framework. The use of speaker-adapted feature-level Bayesian fusion reduced the error rate by 19.4% as compared to the conventional single feature-based confidence scoring in an isolated word out-of-vocabulary rejection test. The decision-level Bayesian fusion also showed better performance than the majority rule. Finally, hybrid Bayesian fusion, which can combine both confidence measure features and local decisions, achieved the best performance.
KW - Adaptive confidence scoring
KW - Bayesian fusion
KW - Confidence measure (CM)
KW - Speech recognition
UR - http://www.scopus.com/inward/record.url?scp=29144490102&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=29144490102&partnerID=8YFLogxK
U2 - 10.1109/LSP.2005.859494
DO - 10.1109/LSP.2005.859494
M3 - Article
AN - SCOPUS:29144490102
VL - 12
SP - 871
EP - 874
JO - IEEE Signal Processing Letters
JF - IEEE Signal Processing Letters
SN - 1070-9908
IS - 12
ER -