TY - GEN
T1 - GMM-based matching ability measurement of a speech recognizer and a feature set
AU - Kim, Hong Kook
AU - Choi, Seung Ho
PY - 2012
Y1 - 2012
N2 - In this work, we propose a Gaussian mixture model-based recognizer selection method to overcome the acoustic mismatch between training and testing environments of a speech recognition system. The method evaluates the preference of a system over other for a specific feature set. By applying it to compare the two speech recognition systems constructed with wireline speech and wireless speech, respectively, it is shown that the matched condition of wireless training and testing can give better recognition accuracies than the mismatched condition.
AB - In this work, we propose a Gaussian mixture model-based recognizer selection method to overcome the acoustic mismatch between training and testing environments of a speech recognition system. The method evaluates the preference of a system over other for a specific feature set. By applying it to compare the two speech recognition systems constructed with wireline speech and wireless speech, respectively, it is shown that the matched condition of wireless training and testing can give better recognition accuracies than the mismatched condition.
KW - acoustic mismatch
KW - Gaussian mixture model
KW - recognizer selection
UR - https://www.scopus.com/pages/publications/84863126163
U2 - 10.1007/978-3-642-27314-8_51
DO - 10.1007/978-3-642-27314-8_51
M3 - Conference contribution
AN - SCOPUS:84863126163
SN - 9783642273131
T3 - Lecture Notes in Electrical Engineering
SP - 377
EP - 383
BT - Future Communication, Computing, Control and Management
T2 - 2011 International Conference on Future Communication, Computing, Control and Management, ICF4C 2011
Y2 - 16 December 2011 through 17 December 2011
ER -