Automatic Pronunciation Evaluation of Non-Native Mandarin Tone by Using Multi-Level Confidence Measures
نویسندگان
چکیده
Automatic evaluation of tonal production plays an important role in a tonal language Computer-Assisted Pronunciation Training (CAPT) system. In this paper, we propose an automatic evaluation method for non-native Mandarin tones. The method applied multi-level confidence measures generated from Deep Neural Network (DNN). The confidence measures consisted of Log Posterior Ratios (LPR), Average Frame-level Log Posteriors (AFLP) and Segment-level Log Posteriors (SLP). The LPR was calculated between the correct tone model and competing tone models. The AFLP and LPR were obtained from frame-level scores. And the SLP was directly derived from segment-level scores. The multi-level confidence measures were modeled with a support vector machine (SVM) classifier. For comparison, three experiments were conducted according to different features: AFLP+LPR, SLP only and AFLP+LPR+SLP. The experimental results showed that the performance of the system which used multilevel confidence measures was the best, achieving a FRR of 5.63% and a DA of 82.45%, which demonstrated the efficiency of the proposed method.
منابع مشابه
Development of a Computer-Aided Language Learning System for Mandarin – Tone Recognition and Pronunciation Error Detection
This paper reports on the continued activities towards the development of a computer-aided language learning system for teaching Mandarin to Germans. A method for f0 normalization based on maximum likelihood estimation and tone recognition was implemented. Furthermore, a method for detecting the pronunciation errors was tested by calculating the confidence distance between the first and second ...
متن کاملTowards Automatic Tone Correction in Non-native Mandarin
Feedback is an important part of foreign language learning and Computer Aided Language Learning (CALL) systems. For pronunciation tutoring, one method to provide feedback is to provide examples of correct speech for the student to imitate. However, this may be frustrating if a student is unable to completely match the example speech. This research advances towards providing feedback using a stu...
متن کاملiCALL corpus: Mandarin Chinese spoken by non-native speakers of European descent
We present iCALL, a speech corpus designed to evaluate Mandarin Chinese pronunciation patterns of non-native speakers of European descent, developed at the Institute for Infocomm Research (IR) in Singapore. To the best of our knowledge, iCALL is larger than any reported non-native corpora to date in terms of utterance number, duration, and number of speakers: iCALL consists of 90,841 utterances...
متن کاملUsing Human Perception for Automatic Accent Assessment
In this study, a new algorithm for automatic accent evaluation of native and non-native speakers is presented. The proposed system consists of two main steps: alignment and scoring. At the alignment step, the speech utterance is processed using a Weighted Finite State Transducer (WFST) based technique to automatically estimate the pronunciation errors. Subsequently, in the scoring step a Maximu...
متن کاملGoodness of tone (GOT) for non-native Mandarin tone recognition
Lexical tone is one of the most challenging pronunciation problems in tonal language acquisition. Accurate lexical tone production is especially challenging for people whose native language is not a tonal one. In this paper, we propose Goodness of Tone (GOT), a confidence measure inspired from goodness of pronunciation (GOP) for tone recognition. GOT is a vector representation of the confidence...
متن کامل