Ιδρυματικό Αποθετήριο [SANDBOX]
Πολυτεχνείο Κρήτης
EN  |  EL

Αναζήτηση

Πλοήγηση

Ο Χώρος μου

Estimation of general identifiable linear dynamic models with an application in speech recognition

Tsontzos Georgios , Diakoloukas Vasilis, Koniaris, Christos, 1979-, Digalakis Vasilis

Απλή Εγγραφή


URIhttp://purl.tuc.gr/dl/dias/DD5D0269-6830-4FB8-8778-977FC068EAEE-
Αναγνωριστικόhttps://doi.org/10.1109/ICASSP.2007.366947-
Αναγνωριστικόhttp://ieeexplore.ieee.org/abstract/document/4218135/-
Γλώσσαen-
ΤίτλοςEstimation of general identifiable linear dynamic models with an application in speech recognitionen
ΔημιουργόςTsontzos Georgios en
ΔημιουργόςDiakoloukas Vasilisen
ΔημιουργόςΔιακολουκας Βασιλeioςel
ΔημιουργόςKoniaris, Christos, 1979-en
ΔημιουργόςDigalakis Vasilisen
ΔημιουργόςΔιγαλακης Βασιληςel
ΕκδότηςInstitute of Electrical and Electronics Engineersen
ΠερίληψηAlthough hidden Markov models (HMMs) provide a relatively efficient modeling framework for speech recognition, they suffer from several shortcomings which set upper bounds in the performance that can be achieved. Alternatively, linear dynamic models (LDM) can be used to model speech segments. Several implementations of LDM have been proposed in the literature. However, all had a restricted structure to satisfy identifiability constraints. In this paper, we relax all these constraints and use a general, canonical form for a linear state-space system that guarantees identifiability for arbitrary state and observation vector dimensions. For this system, we present a novel, element-wise maximum likelihood (ML) estimation method. Classification experiments on the AURORA2 speech database show performance gains compared to HMMs, particularly on highly noisy conditions. en
ΤύποςΔημοσίευση σε Συνέδριοel
ΤύποςConference Publicationen
Άδεια Χρήσηςhttp://creativecommons.org/licenses/by/4.0/en
Ημερομηνία2015-11-08-
Ημερομηνία Δημοσίευσης2007-
Θεματική ΚατηγορίαHMMs (Hidden Markov models)en
Θεματική Κατηγορίαhidden markov modelsen
Θεματική Κατηγορίαhmms hidden markov modelsen
Θεματική ΚατηγορίαMechanical speech recognizeren
Θεματική ΚατηγορίαSpeech recognition, Automaticen
Θεματική Κατηγορίαautomatic speech recognitionen
Θεματική Κατηγορίαmechanical speech recognizeren
Θεματική Κατηγορίαspeech recognition automaticen
Βιβλιογραφική ΑναφοράG. Tsontzos, V. Diakoloukas, C. Koniaris and V. Digalakis, "Estimation of general identifiable linear dynamic models with an application in speech recognition," in 2007 IEEE Int. Conf. on Acoust., Speech and Sign. Process. (ICASSP) doi: 10.1109/ICASSP.2007.366947en

Υπηρεσίες

Στατιστικά