A Hidden Trajectory Model with Bi-directional Target-Filtering: Cascaded vs. Integrated Implementation for Phonetic Recognition

Author

Deng, Li ; Li, Xiang ; Yu, Dong ; Acero, Alex

Author_Institution

Microsoft Res., Redmond, WA, USA

Volume

1

fYear

2005

fDate

March 18-23, 2005

Firstpage

337

Lastpage

340

Keywords

Gaussian distribution; cepstral analysis; error statistics; sequences; series (mathematics); speech processing; speech recognition; Gaussian distribution; HTM; TIMIT phonetic recognition task; VTR dynamics; VTR target filtering; acoustic likelihood; acoustic speech model; bi-directional target filtering; bi-directional vocal tract resonance; cepstrum-based acoustic observation; first-order Taylor series approximation; hidden trajectory model; integrated likelihood; large-scale N-best rescoring; local slope; nonlinear mapping function; phone sequence; phonetic recognition; speech recognition; statistical distribution; statistical hidden trajectory modeling; structured long-span context dependence; time-varying mean; training; two-stage cascading; upper error bound; Bidirectional control; Cepstral analysis; Filtering; Hidden Markov models; Nonlinear acoustics; Resonance; Speech recognition; Target recognition; Trajectory; Video recording;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech, and Signal Processing, 2005. Proceedings. (ICASSP '05). IEEE International Conference on

ISSN

1520-6149

Print_ISBN

0-7803-8874-7

Type

conf

DOI

10.1109/ICASSP.2005.1415119

Filename

1415119