Pseudo 2-dimensional hidden Markov models in speech recognition

S. Werner; G. Rigoll

doi:10.1109/ASRU.2001.1034679

Pseudo 2-dimensional hidden Markov models in speech recognition

S. Werner, G. Rigoll

University of Duisburg-Essen

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

4 Scopus citations

Abstract

In this paper, the usage of pseudo 2-dimensional hidden Markov models for speech recognition is discussed. This image processing method should better model the time-frequency structure in speech signals. The method calculates the emission probability of a standard HMM by embedded HMM for each state. If a temporal sequence of spectral vectors is imagined as a spectrogram, this leads to a 2-dimensional warping of the spectrogram. This additional warping of the frequency axis could be useful for speaker-independent recognition and can be considered to be similar to a vocal tract normalization. The effects of this paradigm are investigated in this paper using the TI-Digits database.

Original language	English
Title of host publication	2001 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2001 - Conference Proceedings
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	441-444
Number of pages	4
ISBN (Electronic)	078037343X, 9780780373433
DOIs	https://doi.org/10.1109/ASRU.2001.1034679
State	Published - 2001
Externally published	Yes
Event	IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2001 - Madonna di Campiglio, Italy Duration: 9 Dec 2001 → 13 Dec 2001

Publication series

Name	2001 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2001 - Conference Proceedings

Conference

Conference	IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2001
Country/Territory	Italy
City	Madonna di Campiglio
Period	9/12/01 → 13/12/01

Access to Document

10.1109/ASRU.2001.1034679

Cite this

Werner, S., & Rigoll, G. (2001). Pseudo 2-dimensional hidden Markov models in speech recognition. In 2001 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2001 - Conference Proceedings (pp. 441-444). Article 1034679 (2001 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2001 - Conference Proceedings). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ASRU.2001.1034679

Werner, S. ; Rigoll, G. / Pseudo 2-dimensional hidden Markov models in speech recognition. 2001 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2001 - Conference Proceedings. Institute of Electrical and Electronics Engineers Inc., 2001. pp. 441-444 (2001 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2001 - Conference Proceedings).

@inproceedings{a499c935a19846678129b4f3fb63f361,

title = "Pseudo 2-dimensional hidden Markov models in speech recognition",

abstract = "In this paper, the usage of pseudo 2-dimensional hidden Markov models for speech recognition is discussed. This image processing method should better model the time-frequency structure in speech signals. The method calculates the emission probability of a standard HMM by embedded HMM for each state. If a temporal sequence of spectral vectors is imagined as a spectrogram, this leads to a 2-dimensional warping of the spectrogram. This additional warping of the frequency axis could be useful for speaker-independent recognition and can be considered to be similar to a vocal tract normalization. The effects of this paradigm are investigated in this paper using the TI-Digits database.",

author = "S. Werner and G. Rigoll",

note = "Publisher Copyright: {\textcopyright} 2001 IEEE.; IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2001 ; Conference date: 09-12-2001 Through 13-12-2001",

year = "2001",

doi = "10.1109/ASRU.2001.1034679",

language = "English",

series = "2001 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2001 - Conference Proceedings",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "441--444",

booktitle = "2001 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2001 - Conference Proceedings",

}

Werner, S & Rigoll, G 2001, Pseudo 2-dimensional hidden Markov models in speech recognition. in 2001 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2001 - Conference Proceedings., 1034679, 2001 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2001 - Conference Proceedings, Institute of Electrical and Electronics Engineers Inc., pp. 441-444, IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2001, Madonna di Campiglio, Italy, 9/12/01. https://doi.org/10.1109/ASRU.2001.1034679

Pseudo 2-dimensional hidden Markov models in speech recognition. / Werner, S.; Rigoll, G.
2001 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2001 - Conference Proceedings. Institute of Electrical and Electronics Engineers Inc., 2001. p. 441-444 1034679 (2001 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2001 - Conference Proceedings).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Pseudo 2-dimensional hidden Markov models in speech recognition

AU - Werner, S.

AU - Rigoll, G.

PY - 2001

Y1 - 2001

N2 - In this paper, the usage of pseudo 2-dimensional hidden Markov models for speech recognition is discussed. This image processing method should better model the time-frequency structure in speech signals. The method calculates the emission probability of a standard HMM by embedded HMM for each state. If a temporal sequence of spectral vectors is imagined as a spectrogram, this leads to a 2-dimensional warping of the spectrogram. This additional warping of the frequency axis could be useful for speaker-independent recognition and can be considered to be similar to a vocal tract normalization. The effects of this paradigm are investigated in this paper using the TI-Digits database.

AB - In this paper, the usage of pseudo 2-dimensional hidden Markov models for speech recognition is discussed. This image processing method should better model the time-frequency structure in speech signals. The method calculates the emission probability of a standard HMM by embedded HMM for each state. If a temporal sequence of spectral vectors is imagined as a spectrogram, this leads to a 2-dimensional warping of the spectrogram. This additional warping of the frequency axis could be useful for speaker-independent recognition and can be considered to be similar to a vocal tract normalization. The effects of this paradigm are investigated in this paper using the TI-Digits database.

UR - http://www.scopus.com/inward/record.url?scp=84962820678&partnerID=8YFLogxK

U2 - 10.1109/ASRU.2001.1034679

DO - 10.1109/ASRU.2001.1034679

M3 - Conference contribution

AN - SCOPUS:84962820678

T3 - 2001 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2001 - Conference Proceedings

SP - 441

EP - 444

BT - 2001 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2001 - Conference Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2001

Y2 - 9 December 2001 through 13 December 2001

ER -

Werner S, Rigoll G. Pseudo 2-dimensional hidden Markov models in speech recognition. In 2001 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2001 - Conference Proceedings. Institute of Electrical and Electronics Engineers Inc. 2001. p. 441-444. 1034679. (2001 IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU 2001 - Conference Proceedings). doi: 10.1109/ASRU.2001.1034679

Pseudo 2-dimensional hidden Markov models in speech recognition

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this