Digalakis, Vassilis - Department of Electronics and Computer Engineering, Technical University of Crete

Advanced Search

Digalakis, Vassilis - Department of Electronics and Computer Engineering, Technical University of Crete

The Stochastic Segment Model for Continuous Speech Recognition M. Ostendorf V. Digalakis
Genones: Generalized Mixture Tying in Continuous Hidden Markov ModelBased Speech Recognizers
In previous work, we showed how to constrain the estimation of continuous mixturedensity hidden Markov models
Speaker Adaptation Using Constrained Estimation of Gaussian Mixtures
COMBINING KNOWLEDGE SOURCES TO REORDER NBEST SPEECH HYPOTHESIS LISTS
DARPA Workshop on Speech and Natural Language, pp. 264269, February 1991.
From HMMs to Segment Models: A Unified View of Stochastic Modeling for Speech Recognition
For the AudioTechnica recordings, there was a 35% increase in the worderror rate when used in a computerroom
FIGURE 1. Average error in the channel estimate (as a percentage of the total variance) as a function of the estimation interval (in seconds) for cepstral coefficients C1 through C8
the Gaussians introduces a significant degradation in recognition performance. This degradation increases when the features are
SRI's DECIPHER TM speech recognition system has been used successfully for largevocabulary continuous speech recognition
T-SA-01047-2005.R1 1 Abstract--The porting of a speech recognition system to a new
Large Vocabulary Continuous Speech Recognition in Greek: Corpus and an Automatic Dictation System
BOSTON UNIVERSITY GRADUATE SCHOOL
ADAPTATION OF HIDDEN MARKOV MODELS USING MULTIPLE STOCHASTIC TRANSFORMATIONS
approximately 10% to approximately 13%). More detail on this work will be presented at the HLT conference and will be included
MAXIMUM LIKELIHOOD IDENTIFICATION OF MULTISCALE STOCHASTIC MODELS USING THE WAVELET TRANSFORM AND THE EM ALGORITHM
The performance and robustness of a speech recognition system can be improved by adapting the speech models to the speaker,
A SPEECH TO SPEECH TRANSLATION SYSTEM BUILT FROM STANDARD COMPONENTS
Speaker Adaptation Using Combined Transformation and Bayesian Methods
Stem-based Maximum Entropy Language Models for Inflectional Languages
DEVELOPMENT OF DIALECTSPECIFIC SPEECH RECOGNIZERS USING ADAPTATION METHODS
The performance and robustness of a speech recognition system can be improved by adapting the speech models to the speaker,
ONLINE ADAPTATION OF HIDDEN MARKOV MODELS USING INCREMENTAL ESTIMATION ALGORITHMS
EFFICIENT SPEECH RECOGNITION USING SUBVECTOR QUANTIZATION AND DISCRETE-MIXTURE HMMS
AUTOMATIC PRONUNCIATION EVALUATION OF FOREIGN SPEAKERS USING UNKNOWN TEXT
A Sub-optimal Viterbi-like Search for Linear Dynamic Models Classification Dimitris Oikonomidis, Vassilis Diakoloukas and Vassilis Digalakis
Degradation of Speech Recognition Performance over Lossy Data Networks
IDENTIFICATION OF LINEAR SYSTEMS IN CANONICAL FORM THROUGH AN EM Pavlos Papadopoulos, Vassilios Digalakis
QUANTIZATION OF CEPSTRAL PARAMETERS FOR SPEECH RECOGNITION OVER THE WORLD WIDE WEB
MAXIMUM LIKELIHOOD STOCHASTIC TRANSFORMATIONS ADAPTATION FOR MEDIUM