Automatic evaluation and training in English pronunciation

Combination of machine scores for automatic grading of pronunciation quality

“…The pronunciation evaluation problem can then be de®ned as an estimation problem, where we try to estimate, or predict, the value of the ideal human grade h by using a set of predictors. These predictors are the machine scores that we have presented in our previous work (Neumeyer et al, 1996(Neumeyer et al, , 1998Rypa, 1996;Bernstein et al, 1990;Bernstein, 1992;Digalakis, 1992;Digalakis and Murveit, 1994), some of which were summarized in the previous section. We investigated the use of linear and nonlinear regression methods to predict one random variable, the human grade, from a set of others, the machine scores.…”

“…The basic pronunciation scoring paradigm (Bernstein et al, 1990;Bernstein, 1992;Digalakis, 1992) uses hidden Markov models (HMMs) (Digalakis and Murveit, 1994) to generate phonetic segmentations of the student's speech. From these segmentations, we use the HMMs to obtain spectral match and duration scores.…”

Avaliação de modelos para reconhecimento automático de fala aplicados para identificação da qualidade de leituras em voz alta de narrativas breves

“…Um dos primeiros trabalhos a utilizar tecnologias de reconhecimento de fala no contexto educacional é o de [Bernstein et al 1990], no qual estudantes japoneses tinham suas habilidades em língua inglesa avaliadas por um sistema que processava áudios de leitura.…”

Assessment of dutch pronunciation by means of automatic speech recognition technology

“…The eventual aim of the research reported on in this paper is to develop an automatic system of pronunciation grading for Dutch by using speech recognition technology. As in other studies [1,2,3], in this investigation the performance of the speech recognizer is validated against pronunciation scores assigned by human experts. Important characteristics of the present investigation as compared to previous ones are that different groups of human experts are involved as raters and that these human raters were required to evaluate several aspects of pronunciation quality.…”