Spoken language pronunciation evaluation method based on deep neural network posterior probability algorithm

Abstract

The present invention discloses a spoken language pronunciation evaluation method based on a deep neural network posterior probability algorithm. The method comprises the following steps of: selectinga certain amount of voice frequencies from voice, wherein the number of words of each voice frequency is in a certain range, calculating the average likelihood of the phoneme of one word, the averageEGOP of the phoneme of one word and the average duration probability of the phoneme of one word in each voice frequency; and taking the average likelihood of the phoneme of one word, the average EGOPof the phoneme of one word and the average duration probability of the phoneme of one word in each voice frequency as input items, inputting the average likelihood of the phoneme of one word, the average EGOP of the phoneme of one word and the average duration probability of the phoneme of one word in each voice frequency into a neural network, and outputting scores of words. The spoken languagepronunciation evaluation method based on a deep neural network posterior probability algorithm starts from an acoustic model, the LSTM modeling is employed to improve the phoneme recognition rate, theFA likelihood and all the similar phoneme likelihoods are compared, a GOP method is extended to an EGOP method, an artificial neural network scoring model is employed to perform scoring so as to obtain an accurate voice evaluation result.

Description

technical field [0001] The invention relates to the field of pronunciation evaluation, in particular to a spoken pronunciation evaluation method based on a deep neural network posterior probability algorithm. Background technique [0002] Commonly used speech evaluation technologies, such as speech evaluation used in oral English teaching, generally use intelligent scoring technology to evaluate learners' spoken English, but the current intelligent scoring technology is mainly based on the GOP (Goodness Of Pronunciation) method. The GOP method relies on two processes, one is Forced Alignment (FA for short), and the other is Free Phoneme (FP for short) decoding, where FA is based on acoustic models and reference texts (that is, learners need to read along Text) find the time boundary of each word, and get the likelihood of each word (Likelihood); while FP decoding uses the same audio, but the unit of decoding is the phoneme level, and each phoneme can be compared with any oth...

AI Technical Summary

Problems solved by technology

However, under normal circumstances, the acoustic model has a relatively low recognition rate for phonemes, so the accuracy of the likelihood obtained by FP decoding is also relatively low, which will lead to inaccurate scoring results.

Images