Title :
Using output probability distribution for oov word rejection
Author :
Huang, Shilei ; Xie, Xiang ; Fung, Pascale
Author_Institution :
Dept. of Electron. & Comput. Eng., Univ. of Sci. & Technol. (HKUST), Hong Kong
Abstract :
This paper proposes a method to calculate the confidence score for out-of-vocabulary (OOV) word verification based on the Output Probability Distribution (OPD) of phoneme HMMs. Compared with input vector for dynamic garbage model, OPD vector contains more information than the sorted probabilities. Confidence score of each phoneme is calculated by SVM with OPD vectors as input. Hypotheses are accepted or rejected based on this confidence score. Experimental results showed that the proposed method achieved lower EER in word verification task than the conventional dynamic garbage model.
Keywords :
hidden Markov models; speech processing; statistical distributions; confidence score; dynamic garbage model; out-of-vocabulary word verification; output probability distribution; phoneme hidden Markov model; Acoustic devices; Automatic speech recognition; Context modeling; Hidden Markov models; Humans; Length measurement; Probability distribution; Speech recognition; Support vector machine classification; Support vector machines; confidence measure; speech recognition; word verification;
Conference_Titel :
Spoken Language Technology Workshop, 2008. SLT 2008. IEEE
Conference_Location :
Goa
Print_ISBN :
978-1-4244-3471-8
Electronic_ISBN :
978-1-4244-3472-5
DOI :
10.1109/SLT.2008.4777880
