Incorporation of the Time-Scale Modification of Speech Processing System into Learning and Instructional System for Listening Comprehension of a Foreign Language

本文讨论了语音信号时间尺度修改系统(以下简称S处理系统)在教学硬件中的有效应用。展示了S处理系统在教学环境中的应用。考察了新的S处理系统在同步视觉呈现中的效果。改进后的系统是对语音信号的非线性时间尺度修改，其重点放在语音和非语音部分。为了弄清发音率和口语单元长度对外语听力理解的影响，实验表明，S处理系统的发音速度变慢，在语音感知研究的基础上，对外语语音识别过程进行了研究，测量了正常和S处理后的扩展语音的感知单位，并进行了间隙检测，测量了间隙检测阈值。本文提出了一个外语语音形式的习得模型和一个词汇通达模型。

Effective incorporation of a system for the time-scale modification of speech signals (henceforth, S-processing) into educational hardware has been discussed.The application of S-processing in the educational setting has been demonstrated.The effect of the new S-processing system when applied to the synchronized visual presentation has been examined.The improved system is nonlinear time-scale modification of speech signals with its focus on the speech and non-speech parts of utterances.The experiment carried with the aim of clarifying the effect of articulation rate and length of spoken unit on foreign language listening comprehension has shown that the slowing of the articulation rate with the S-Processing System, and the insertion of pauses, with the Voice Analyzing System which shortens the spoken unit, facilitated listening comprehension.Increasing pauses yielded more significant improvement in comprehension than decreasing the articulation rate.Processes of foreign language speech recognition have been considered on the basis of studies on speech perception.Perceptual sense unit has been measured both in normal and s-processed expanded speeches.Gap detection tests have been developed and gap detection threshold has been measured. They have turned out to be an accurate quantitative measure of listening ability.An acquisition model of a phonetic form in a foreign language and a lexical access model have been proposed.