Title | ||
---|---|---|
Dynamic Bayesian Network Based Speech Recognition with Pitch and Energy as Auxiliary Variables |
Abstract | ||
---|---|---|
Abstract: Pitch and energy are two fundamental features describingspeech, having importance in human speech recognition.However, when incorporated as features in automatic speech recognition(ASR), they usually result in a signicant degradation onrecognition performance due to the noise inherent in estimating ormodeling them. In this paper, we show experimentally how thiscan be corrected by either conditioning the emission distributionsupon these features or by marginalizing out these features in ... |
Year | DOI | Venue |
---|---|---|
2002 | 10.1109/NNSP.2002.1030075 | NNSP |
Keywords | Field | DocType |
belief networks,feature extraction,learning (artificial intelligence),parameter estimation,random noise,speech recognition,HMM,acoustic feature estimation,automatic speech recognition,dynamic Bayesian networks,emission distributions,energy,hidden Markov models,pitch,training data | Pattern recognition,Computer science,Random noise,Feature extraction,Auxiliary variables,Speech recognition,Speaker recognition,Artificial intelligence,Estimation theory,Hidden Markov model,Dynamic Bayesian network | Conference |
Citations | PageRank | References |
7 | 0.64 | 6 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Todd A. Stephenson | 1 | 82 | 6.12 |
Jaume Escofet | 2 | 7 | 0.64 |
Mathew Magimai-Doss | 3 | 516 | 54.76 |
Herve Bourlard | 4 | 152 | 37.75 |