Posterior Probability Based Confidence Measures Applied to a Children's Speech Reading Tracking System

نویسندگان

  • Daniel Bolaños
  • Wayne H. Ward
چکیده

In this paper, we present improved wordlevel confidence measures based on posterior probabilities for children’s oral reading continuous speech recognition. Initially we compute posterior probability based confidence measures on word graphs using a forward-backward algorithm. We study how an increase of the word graph density affects the quality of these confidence measures. For this purpose we merge word graphs obtained using three different language models and compute the previous confidence measures over the resulting word graph. This produces a relative error reduction of 8% in Confidence Error Rate compared to the baseline confidence measure. Moreover the system operating range is increased significantly.

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

On Using Entropy Information to Improve Posterior Probability-Based Confidence Measures

In this paper, we propose a novel approach that reduces the confidence error rate of traditional posterior probability-based confidence measures in large vocabulary continuous speech recognition systems. The method enhances the discriminability of confidence measures by applying entropy information to the posterior probability-based confidence measures of word hypotheses. The experiments conduc...

متن کامل

Confidence measures for phonetic segmentation of continuous speech

In the context of text-to-speech synthesis, this contribution deals with the segmentation of speech into phone units. Using an HMM based segmentation system, we proceed to compare several phone-level confidence measures to detect potential local mismatches between the phone labels and the acoustics. As well as serving this purpose, these confidence measures will help the system suggest a new lo...

متن کامل

Confidence measures for hybrid HMM/ANN speech recognition

In this paper we introduce four acoustic confidence measures which are derived from the output of a hybrid HMM/ANN large vocabulary continuous speech recognition system. These confidence measures, based on local posterior probability estimates computed by an ANN, are evaluated at both phone and word levels, using the North American Business News corpus.

متن کامل

Confidence Measure Based Model Adaptation for Speaker Verification

Confidence measures are expected give a measure of reliability on the result of a speech/speaker recognition system. Most commonly used confidence measures are based on posterior word or phoneme probabilities which can be obtained from the output of the recognizer. In this paper we introduced a linear interpretation of posterior probability based confidence measure by using inverse Fisher trans...

متن کامل

Speaker Model Adaptation Based on Confidence Score

Original scientific paper Confidence measures are expected to give a measure of reliability on the result of a speech/speaker recognition system. Most commonly used confidence measures are based on posterior word or phoneme probabilities which can be obtained from the output of the recognizer. In this paper we introduced a linear interpretation of posterior probability based confidence measure ...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2007