KR950034012A

KR950034012A - Language training system based on language synthesis

Info

Publication number: KR950034012A
Application number: KR1019950005079A
Authority: KR
Inventors: 라울 자브킨 헥토르; 그레이스 키트 엘리자베스; 안토난자스-바로소 노르마; 알렌 한슨 브리안
Original assignee: 도모유끼 다끼모또; 마쯔시따 일렉트릭 인더스트리얼 코퍼레이션 리미티드
Priority date: 1994-04-12
Filing date: 1995-03-13
Publication date: 1995-12-26
Also published as: CN1115900A; CN1064766C; KR0149622B1

Abstract

본 발명은 학생이 학습한 발음을 입력하고 발음을 내기 위해 필요한 유성음 모델운동이 CRT스트린상에 표시되도록 하는 언어 훈련 시스템에 관한 것이다. 본 시스템은 타자된 발음을 받아들여 음소 또는 음절들일 수 있는 스피치 유닛들의 세트 및 스피치 유닛들의 시점 및 종점으로 분해한다. 스피치 유닛들의 세트는 합성장치로 보내려 발음의 음향학적 특성을 나타내는 변수들의 세트로 바뀐다. 음향학적 변수들은 타자된 발음을 만들기 위해 필요한 주파수 및 비음도를 강조하는 유성음 변수들로 변환된다.The present invention relates to a language training system in which a voiced model movement necessary for inputting and producing pronunciations that a student has learned is displayed on a CRT screen. The system accepts the typed pronunciation and breaks it down into a set of speech units, which can be phonemes or syllables, and the start and end points of speech units. The set of speech units is converted to a set of variables that are sent to the synthesizer to represent the acoustic characteristics of the pronunciation. Acoustic variables are converted into voiced voice variables that emphasize the frequency and non-phonographies needed to make typed pronunciation.

스피치 유닛들 및 각 스피치 유닛의 시점 및 종점은 타자된 발음을 만들기 위해 필요한 혀-구개 접촉형태들을 형성하기 위해 사용된다. 유성음 변수들은 또한 포먼트 합성장치로 보내져 스피치 출력으로 바뀐다. 시스템은 학생의 성과를 측정하여 유사성을 위해, 타자된 발음의 변수들에 대한 학생의 성과를 평가한다. 유사성에 대한 피드백이 CRT스크린상에 표시된다.Speech units and the start and end points of each speech unit are used to form the tongue- palate contacts required to make the typed pronunciation. Voiced variables are also sent to the formant synthesizer to convert to speech output. The system measures student performance and evaluates student performance for typed pronunciation variables for similarity. Feedback on similarity is displayed on the CRT screen.

Description

Language training system based on language synthesis

본 내용은 요부공개 건이므로 전문내용을 수록하지 않았음Since this is an open matter, no full text was included.

제1도는 단어에 대한 아스키(ASCII) 철자표현을 음소들, 강세 및 언어요소 정보로 바꾸는데 있어서 포함되는 단계들을 나타내는 블록 다이어그램이다, 제2도는 언어 훈련 시스템을 나타내는 블록다이어그램이다, 제3도는 발음시 생기는 무비음 및 극비음의 음향학적 변수들을 나타내는 음성스펙트럼이다, 제4도는 발음시 입천정에 대한 혀접촉 방식의 변수들을 나타내는 음성스펙트럼이다, 제5도는 발음시 생기는 실제의 혀접촉 방식을 나타낸다.FIG. 1 is a block diagram showing the steps involved in converting ASCII spellings of words into phonemes, accents, and language element information. FIG. 2 is a block diagram showing a language training system. FIG. Fig. 4 is a voice spectrum representing the variable of the tongue contact method for the palate during pronunciation. Fig. 5 shows the actual tongue contact method during pronunciation.

Claims

A language training apparatus based on language synthesis, comprising: an input device for inputting a pronunciation to be expressed, a set of speech units and an editing device for converting the pronunciation to a viewpoint and a midpoint of each speech unit, a set of speech units And a synthesizing apparatus for converting the start and end points of each speech unit into a set of tongue palatal contact forms and a device for displaying a set of tongue palatal contact forms.

2. The language training apparatus of claim 1, wherein the synthesizing apparatus provides duration of portions of the start, end and contact steady state of the speech unit.

3. The language training apparatus of claim 2, wherein the starting point, the end point, and the steady state viewpoints are allocated for each speech unit by the synthesizing apparatus.

4. The language training apparatus of claim 3, wherein the synthesizing apparatus stores the maximum area of tongue- palate contacts for each speech unit with at least one alternating contact surface for each different context in which the speech units are used.

5. The language training apparatus of claim 4, wherein each contact surface between the tongue and the palate is defined by one set of 63 points.

2. The language training apparatus of claim 1, further comprising a device for measuring the pronunciation produced, including measuring the tongue- palatal contact pattern.

7. The language training system of claim 6, further comprising an apparatus for evaluating the similarity between the pronunciation to be expressed and the measurement of tongue- palate contact form.

8. The language training apparatus of claim 7, further comprising a device for displaying an evaluation result of the similarity between the pronunciation to be expressed and the measured tongue- palate contact form.

A language training method based on linguistic synthesis, the method comprising: inputting a pronunciation to be expressed, converting the pronunciation into a speech unit and the start and end points of the speech unit, and converting the speech unit and the start and end points into a set of tongue-pallate contacts. And converting and displaying a set of tongue- palate contacts.

10. The method of claim 9 wherein converting speech units comprises storing a tongue- palate contact area for each speech unit and allocating time points, endpoints and phase states for each speech unit. .

The method of claim 10, wherein storing the tongue palatal contact area comprises storing a maximum tongue palatal contact area with at least one alternating contact surface for each different context in which the speech unit is used. .

12. The method of claim 11 wherein the storing step comprises storing each contact area between the tongue and the palate by a specific set of 63 dots.

※ Note: The disclosure is based on the initial application.