Speech Recognition : New Techniques for Speaker Adaptation
نویسندگان
چکیده
Résumé : Les systèmes de reconnaissance de la parole utilisant des modèles acoustiques dépendants du locuteur sont plus performants que ceux basés sur des modèles indépendants du locuteur. Le but des techniques d'adaptation est d'améliorer ces derniers modèles pour s'approcher des performances obtenues avec un modéle dépendant du locuteur. Dans cet article, nous proposons deux nouvelles méthodes d'adaptation. La première utilisant les données de test et d'apprentissage pour adapter les modèles indépendants du locuteur, la seconde étant une technique d'adaptation basée sur une classification hiérarchique des gaussiennes composant le modèle acoustique. Ces stratégies d'adaptation ont été évaluées sur le corpus de test de l'AUPELF, ARC B1. Ces deux techniques permettent respectivement un gain relatif par rapport au système initial de 15% pour la première technique et de 16% pour la seconde.
منابع مشابه
Speaker Adaptation in Continuous Speech Recognition Using MLLR-Based MAP Estimation
A variety of methods are used for speaker adaptation in speech recognition. In some techniques, such as MAP estimation, only the models with available training data are updated. Hence, large amounts of training data are required in order to have significant recognition improvements. In some others, such as MLLR, where several general transformations are applied to model clusters, the results ar...
متن کاملSpeaker Adaptation in Continuous Speech Recognition Using MLLR-Based MAP Estimation
A variety of methods are used for speaker adaptation in speech recognition. In some techniques, such as MAP estimation, only the models with available training data are updated. Hence, large amounts of training data are required in order to have significant recognition improvements. In some others, such as MLLR, where several general transformations are applied to model clusters, the results ar...
متن کاملThe use of speaker correlation information for automatic speech recognition
This dissertation addresses the independence of observations assumption which is typically made by today’s automatic speech recognition systems. This assumption ignores within-speaker correlations which are known to exist. The assumption clearly damages the recognition ability of standard speaker independent systems, as can seen by the severe drop in performance exhibited by systems between the...
متن کاملACOUSTIC MODEL ADAPTATION FOR AUTOMATIC SPEECH RECOGNITION AND ANIMAL VOCALIZATION CLASSIFICATION by
ACOUSTIC MODEL ADAPTATION FOR AUTOMATIC SPEECH RECOGNITION AND ANIMAL VOCALIZATION CLASSIFICATION Jidong Tao, B.Eng., M.S. Marquette University, 2009 Automatic speech recognition (ASR) converts human speech to readable text. Acoustic model adaptation, also called speaker adaptation, is one of the most promising techniques in ASR for improving recognition accuracy. Adaptation works by tuning a g...
متن کاملEigenvoices for speaker adaptation
We have devised a new class of fast adaptation techniques for speech recognition, based on prior knowledge of speaker variation. To obtain this prior knowledge, one applies Principal Component Analysis (PCA) [9] or a similar technique to a training set of T vectors of dimension D derived from T speaker-dependent (SD) models. This offline step yields T basis vectors, which we call “eigenvoices” ...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2004