Semi-monitoring speaker self-adaption

Abstract

To prevent adaptation to misrecognized words in unsupervised or on-line automatic speech recognition systems confidence measures are used or the user reaction is interpreted to decide whether a recognized phoneme, several phonemes, a word, several words or a whole utterance should be used for adaptation of the speaker independent model set to a speaker adapted model set or not and, in case an adaptation is executed, how strong the adaptation with this recognized utterance or part of this recognized utterance should be performed. Furtheron, a verification of the speaker adaptation performance is proposed to secure that the recognition rate never decreases (significantly), but only increases or stays at the same level.

Description

technical field [0001] The present invention relates to Automatic Speech Recognition (ASR), and in particular to a method of performing unsupervised or on-line adaptation of an Automatic Speech Recognition system and a speech recognition system capable of implementing the method of the invention. Background technique [0002] State-of-the-art speech recognizers include a set of statistical distributions that model the acoustic properties of certain speech segments. These acoustic properties are encoded as feature vectors. As an example, a Gaussian distribution can be made for each syllable. These distributions are hooked to a certain state. A (stochastic) state transition network (usually a Hidden Markov Model) defines probabilities for state sequences and feature vector sequences. Passing through a state requires the use of a feature vector that covers a frame of speech signal with a length of, for example, 10 ms. [0003] The stochastic parameters of such a recognizer ...

AI Technical Summary

Problems solved by technology

When this happens repeatedly, recognition performance drops dramatically

Images