Title :
A chunk-based phonetic score for mobile voice search
Author :
Prabhavalkar, Rohit ; Droppo, Jasha
Author_Institution :
Dept. of Comput. Sci. & Eng., Ohio State Univ., Columbus, OH, USA
Abstract :
We propose a chunk-based phonetic score for re-scoring word hypotheses for the mobile voice search task. The score is based on a novel technique for aligning decoded phone sequences with forced-alignments of hypothesized word sequences and exploits phone-boundary timing information. In experimental results, we find that the proposed approach results in relative a word error rate reduction of 4.4% and a relative sentence error rate reduction of 2.3% for the Windows live search for mobile task [1].
Keywords :
audio signal processing; error statistics; information retrieval systems; speech processing; Windows live search; chunk based phonetic score; forced alignments; hypothesized word sequences; mobile voice search task; phone boundary timing information; phone sequences; pronounciation modeling; relative sentence error rate reduction; word error rate reduction; word hypotheses; Acoustics; Computational modeling; Decoding; Equations; Hidden Markov models; Mathematical model; Mobile communication; phonetic score; pronunciation modeling; voice search;
Conference_Titel :
Acoustics, Speech and Signal Processing (ICASSP), 2012 IEEE International Conference on
Conference_Location :
Kyoto
Print_ISBN :
978-1-4673-0045-2
Electronic_ISBN :
1520-6149
DOI :
10.1109/ICASSP.2012.6288975
