Base on our former experience in the “ITRI 104 Auto Attendant System”[1] of using keyword spotting for Mandarin speech recognition, a Chinese-English mixedlingual keyword spotting for catering the speaking style of Taiwanese is present. Detailed descriptions and discussions for developing the mixed-lingual autoattendant system are included in this paper, especially for solving different scoring scales in the decoding phase and the re-scoring phase for these two languages. In the decoding phase, we propose a bias-compensation method to make up the score-gap in the likelihood calculation of using Chinese and English acoustic models. To select the most probable result from the recognized hypotheses, the method for normalizing the combination scores of using different scoring mechanisms in the re-scoring phase is also presented.
Cite as: You, S., Chien, S., Hsu, C., Chen, K., Tu, J., Lin, J.S., Chang, S. (2004) Chinese-English Mixed-Lingual Keyword Spotting. Proc. International Symposium on Chinese Spoken Language Processing, 237-240
@inproceedings{you04_iscslp,
author={ShanRuei You and ShihChieh Chien and ChihHsing Hsu and KeShiu Chen and JiaJang Tu and Jeng Shien Lin and SenChia Chang},
title={{Chinese-English Mixed-Lingual Keyword Spotting}},
year=2004,
booktitle={Proc. International Symposium on Chinese Spoken Language Processing},
pages={237--240}
}