| 摘要 |
A user interface 200, a system and a method are provided to automatically compare the speech signal of a language learner 220 against that of a language teacher 210. The system labels the input speech with phonic symbols 216 and identifies the portions where the difference is significant. The system then gives grades and suggestions to the learners for improvement. The comparisons and suggestions include articulation correctness, timing, pitch, intensity, etc.. The method comprises three major stages. In the first stage, a phoneme-feature database is established which contains the statistical data of phonemes. In the second stage, the speech signals of a language learner and a language teacher are labelled with phonic symbols that represent phonemes. In the third stage, the corresponding sections in the student and teacher's speech signals are identified and compared. Grades and suggestions for improvement are given. The user interface for automatically labelling phonic symbols comprises, for each of the two sound signals, a waveform graph, an intensity variation graph obtained by analysing the sound signal, a pitch variation graph and a plurality of pronunciation intervals, wherein each interval comprises a plurality of neighbouring frames attributed to the same phonic cluster and a phonic symbol labelling area. A system is also disclosed which comprises an input device, an electronic phonetic dictionary, an audio cutter as well as the phoneme feature database and phonic symbol labelling device. |
