Welcome to the InfoSci Platform
Feedback-Driven Refinement of Mandarin Speech Recognition Result based on Lattice Modification and Rescoring

Feedback-Driven Refinement of Mandarin Speech Recognition Result based on Lattice Modification and Rescoring

Xiangdong Wang, Yang Yang, Hong Liu, Yueliang Qian, Duan Jia
Copyright: © 2017 |Volume: 9 |Issue: 2 |Pages: 10
ISSN: 1937-965X|EISSN: 1937-9668|EISBN13: 9781522512080|DOI: 10.4018/IJAPUC.2017040104
Cite Article Cite Article

MLA

Wang, Xiangdong, et al. "Feedback-Driven Refinement of Mandarin Speech Recognition Result based on Lattice Modification and Rescoring." IJAPUC vol.9, no.2 2017: pp.55-64. http://doi.org/10.4018/IJAPUC.2017040104

APA

Wang, X., Yang, Y., Liu, H., Qian, Y., & Jia, D. (2017). Feedback-Driven Refinement of Mandarin Speech Recognition Result based on Lattice Modification and Rescoring. International Journal of Advanced Pervasive and Ubiquitous Computing (IJAPUC), 9(2), 55-64. http://doi.org/10.4018/IJAPUC.2017040104

Chicago

Wang, Xiangdong, et al. "Feedback-Driven Refinement of Mandarin Speech Recognition Result based on Lattice Modification and Rescoring," International Journal of Advanced Pervasive and Ubiquitous Computing (IJAPUC) 9, no.2: 55-64. http://doi.org/10.4018/IJAPUC.2017040104

Export Reference

Mendeley
Favorite Full-Issue Download

Abstract

In real world applications of speech recognition, recognition errors are inevitable, and manual correction is necessary. This paper presents an approach for the refinement of Mandarin speech recognition result by exploiting user feedback. An interface incorporating character-based candidate lists and feedback-driven updating of the candidate lists is introduced. For dynamic updating of candidate lists, a novel method based on lattice modification and rescoring is proposed. By adding words with similar pronunciations to the candidates next to the corrected character into the lattice and then performing rescoring on the modified lattice, the proposed method can improve the accuracy of the candidate lists even if the correct characters are not in the original lattice, with much lower computational cost than that of the speech re-recognition methods. Experimental results show that the proposed method can reduce 24.03% of user inputs and improve average candidate rank by 25.31%.

Request Access

You do not own this content. Please login to recommend this title to your institution's librarian or purchase it from the IGI Global bookstore.