Spoken term detection for OOV terms based on triphone confusion matrix

Author

Yong Xu ; Wu Guo ; Shan Su ; Lirong Dai

Author_Institution

Dept. of Electron. Eng. & Inf. Sci., Univ. of Sci. & Technol. of China, Hefei, China

fYear

2012

fDate

5-8 Dec. 2012

Firstpage

98

Lastpage

102

Abstract

The search for out of vocabulary (OOV) query terms in spoken term detection (STD) task is addressed in this paper. The phone level fragment with word-position marker is naturally adopted as the speech recognition decoding unit. Then the triphone confusion matrix (TriCM) is used to expand the query space to compensate for speech recognition errors. And we also propose a new approach to construct triphone confusion matrix using a smoothing method similar with the Katz method to solve the data sparseness problem. Experimental result on the NIST STD06 eval-set conversational telephone speech (CTS) corpus indicates that triphone confusion matrix can provide a relative improvement of 12% in actual term weighted value (ATWV).

Keywords

decoding; query processing; signal detection; smoothing methods; speech coding; speech recognition; telephone sets; vocabulary; ATWV; CTS corpus; Katz method; NIST STD06 eval-set conversational telephone speech; OOV query term; STD task; TriCM; actual term weighted value; data sparseness problem; out of vocabulary; phone level fragment; query space; smoothing method; speech recognition decoding unit; speech recognition error; spoken term detection; triphone confusion matrix; word-position marker; Decoding; Hidden Markov models; Indexes; NIST; Speech; Speech recognition; Vocabulary; out of vocabulary; positioned fragment; spoken term detection; triphone confusion matrix;

fLanguage

English

Publisher

ieee

Conference_Titel

Chinese Spoken Language Processing (ISCSLP), 2012 8th International Symposium on

Conference_Location

Kowloon

Print_ISBN

978-1-4673-2506-6

Electronic_ISBN

978-1-4673-2505-9

Type

conf

DOI

10.1109/ISCSLP.2012.6423480

Filename

6423480