JP2004240352A

JP2004240352A - Speech comparison learning device

Info

Publication number: JP2004240352A
Application number: JP2003031955A
Authority: JP
Inventors: Yukitoshi Cho; 志敏張
Original assignee: Individual
Current assignee: Individual
Priority date: 2003-02-10
Filing date: 2003-02-10
Publication date: 2004-08-26

Abstract

<P>PROBLEM TO BE SOLVED: To make it possible to easily comparatively learn speech. <P>SOLUTION: This learning device is so configured as to be able to repeat the features of " rewind, fast forward, playback, and reproducing speed change" only by IC memory when combining an ordinary speech medium with IC memory and once recording the speech data from the medium into the IC memory. Then, the learning device permits to re-confirm sample speech for a textbook and own speech by own ear by reproducing them from the IC memory, and thereby permits comparison function between own speech and the sample speech for the textbook. Moreover, the learning device is so arranged as to be able to change playback speed of the speech by changing the speed for accessing to the speech data in the IC memory. <P>COPYRIGHT: (C)2004,JPO&NCIPI

Description

【０００１】
【発明の属する技術分野】音声学習技術分野
【０００２】
【従来の技術】従来の技術ではメディアが音声を再生と記録できるが、二つ音声の比較再生ができない、或いはコストが高い、操作が煩わしく普及できない。本発明はこれらの問題をＩＣメモリの介在で簡単に解決できる。さらにほとんどの操作はＩＣメモリを利用するため、サンプル音声に影響しない、機械的の消耗は最大限に抑える。
【０００３】
【特許文献１】特許出願平１１−２２４６１０
【特許文献２】特許出願平６−１７５０３５
【０００４】
【発明が解決しようとする課題】低コスト実用的な音声比較学習器を作成。
【課題を解決するための手段】普通の音声メディアとＩＣメモリを結合し、メディアからＩＣメモリ中に一度音声データを記録した場合、ＩＣメモリだけで「巻き戻し、早送り、再生、再生速度変更」の機能を繰返し実現できる。教本となるサンプル音声と自分の音声をＩＣメモリの再生により、自分の音声を自分の耳で再確認し、教本となるサンプルの音声と比較機能を実現。
【発明の実施の形態】
・音源再生コントロール▲１▼でサンプル音声（Ａ）を再生するとき、ＩＣメモリに自動的に記録する。アナログの音声をデジタルに変換し、サンプリングとＩ／Ｏ処理及び圧縮処理によって、ＩＣメモリに記録できるわけ。
・メモリに記録されたデジタル音声データを再生するには、再生シグナルをサウンドデータアクセスコントロールに送り、ＩＣメモリのデータ記憶領域ＳＤ（Ａ）からデータを順番に呼び出し、データ復元とＩ／Ｏ処理を経由して、アンプで再生できる。
・繰り返しの再生は機械的なメディアを利用せず、ＩＣメモリで実現できるわけ。
・上記の手順と同様に、人間の音声（Ｂ）もマイク▲３▼でＩＣメモリに記録、再生できる。
・音声（Ａ）または音声（Ｂ）の再生中に、速度コンロールシグナルでデータの呼び出し速度を変えると、再生速度を自由に変更できる。聞き取りにくい短い音声も速度落としてじっくり聞くことができる。またＩＣメモリの音声データ再生を一時停止▲６▼あるいは高速に前後検索▲９▼ができる。
テスト結果により：
速度を遅くする場合、約平常の６６％−−８３％。音声を細かく聞き取れ、分析できる。
速度を速くする場合、約平常の１２０％−−１５０％。ヒアリングの強化訓練ができる。
・事前に学習プログラムを組めば、自動的に学習機能を実現できる。このプログラムのフローチャートは図３にて提示。
・人間は自分が発音するとき、その正確さを把握にくい。自分の声をメディアに記録し、教本となるサンプル音声と比較しないと自分の欠点が分からない。この発明は人間の音声学習課題を解決につながる。
教本となるサンプル音声（Ａ）は再生時、ＩＣメモリに自動的記録される。サンプル音声（Ａ）に一定時間の空白があれば、区切りと認識し、ＩＣメモリ中に記録されたサウンドデータＳＤ（Ａ）が自動的に数回再生。もしＳＤ（Ｂ）に人の声も吹き込んだ場合、ＳＤ（Ａ）に続きとＳＤ（Ｂ）と連続再生を数回繰り返す、ＳＤ（Ａ）とＳＤ（Ｂ）の比較になる。その後、サウンドデータをリセットし、サンプル音声（Ａ）の再生を継続。
【実施例】語学教育、歌と発音などの音声教育、聴覚と発音に障害のある方のトレーニング。
【発明の効果】語学教育、歌と発音などの音声教育、聴覚と発音に障害のある方のトレーニングに役に立つ。
【図面の簡単な説明】
【図１】音声比較学習器の機能説明図
【図２】音声比較学習器の構成原理図
【図３】音声学習機能プログラムのフローチャート
【符号の説明】
ＳＤ：サウンドデータ（ＳｏｕｎｄＤａｔａ）
ＩＣメモリ：データを記録できる集積回路
音声検索：ＩＣメモリに記録した音声データをアクセスし、検索すること
Ｄ＜＝＞Ａ変換：デジタル信号とアナログ信号の双方向変換
ＣＰＵ：セントラルプロセスユニット
Ｉ／Ｏ処理：データの入出力処理[0001]
TECHNICAL FIELD The field of speech learning technology
2. Description of the Related Art In the prior art, a medium can reproduce and record sound, but cannot compare and reproduce two sounds, or the cost is high, the operation is cumbersome, and it cannot be spread. The present invention can easily solve these problems with the intervention of an IC memory. In addition, most operations utilize IC memory, which minimizes mechanical wear which does not affect sampled audio.
[0003]
[Patent Document 1] Japanese Patent Application No. 11-224610
[Patent Document 2] Japanese Patent Application No. 6-175035
[0004]
SUMMARY OF THE INVENTION A low-cost practical sound comparison learning device is created.
Means for Solving the Problems When a normal audio medium and an IC memory are combined and audio data is recorded once from the medium in the IC memory, "rewind, fast forward, reproduction, change of reproduction speed" is performed only by the IC memory. Function can be realized repeatedly. Reproduce your own voice with your own ears by playing the sample voice and your own voice in the IC memory as the textbook, and realize the comparison function with the voice of the sample as the textbook.
BEST MODE FOR CARRYING OUT THE INVENTION
When the sample sound (A) is reproduced by the sound source reproduction control (1), it is automatically recorded in the IC memory. That is, analog sound can be converted to digital and recorded in IC memory by sampling, I / O processing and compression processing.
-To reproduce digital audio data recorded in the memory, a reproduction signal is sent to the sound data access control, data is sequentially called from the data storage area SD (A) of the IC memory, and data restoration and I / O processing are performed. Via, can be played on the amplifier.
-Repetitive reproduction can be realized by IC memory without using mechanical media.
Similarly to the above procedure, the human voice (B) can be recorded and reproduced in the IC memory by the microphone (3).
When the data calling speed is changed by the speed control signal during the reproduction of the voice (A) or the voice (B), the reproduction speed can be freely changed. Even short voices that are difficult to hear can be heard at a reduced speed. In addition, the reproduction of audio data in the IC memory can be temporarily stopped (6) or a forward / backward search (9) can be performed at high speed.
According to the test results:
When reducing the speed, about 66% -83% of normal. Can hear and analyze voices in detail.
When increasing the speed, about 120% -150% of normal. Hearing training can be conducted.
-If a learning program is set up in advance, the learning function can be automatically realized. The flowchart of this program is presented in FIG.
・ It is difficult for humans to understand the accuracy of their pronunciation. If you do not record your voice on the media and compare it to the sample voice used as the textbook, you will not be able to understand your own shortcomings. The present invention solves the problem of human speech learning.
The sample voice (A) serving as a textbook is automatically recorded in the IC memory during reproduction. If there is a blank space in the sample sound (A) for a certain period of time, it is recognized as a break and the sound data SD (A) recorded in the IC memory is automatically reproduced several times. If a human voice is also blown into SD (B), SD (A) and SD (B) will be compared with SD (A) and SD (B) will be repeated several times. After that, the sound data is reset, and the reproduction of the sample sound (A) is continued.
[Example] Language education, voice education such as singing and pronunciation, training for people with hearing and pronunciation impairments.
The present invention is useful for language education, voice education such as singing and pronunciation, and training for those with impaired hearing and pronunciation.
[Brief description of the drawings]
FIG. 1 is a functional explanatory diagram of a speech comparison learning device. FIG. 2 is a configuration diagram of a speech comparison learning device. FIG. 3 is a flowchart of a speech learning function program.
SD: Sound Data (Sound Data)
IC memory: Integrated circuit capable of recording data Voice search: Accessing and searching voice data recorded in IC memory D <=> A conversion: Bidirectional conversion between digital and analog signals CPU: Central process unit I / O Processing: Data input / output processing

Claims

ＩＣメモリ中に一度音声データを記録した場合、ＩＣメモリだけで「巻き戻し、早送り、再生、再生速度変更」の機能を繰返し実現できる。このため、これらの動作をモータなど機械的な要素から最大限に切り離しできる。テープなど音源メディアの走行消耗を最大限に抑え、高速な音声検索と機敏な再生を実現。When audio data is recorded once in the IC memory, the functions of “rewind, fast forward, play, change playback speed” can be realized repeatedly with the IC memory alone. For this reason, these operations can be maximally separated from mechanical elements such as a motor. It minimizes the running wear of sound source media such as tapes, and achieves high-speed voice search and agile playback.

自分の音声を自分の耳で再確認し、教本となるサンプルの音声と比較機能を実現。図１で示すように、テープ・ＣＤ・ＭＤ・ＭＰ３など学習サンプル音源からの音声ＡをＩＣメモリに記録すること（▲１▼）により、メモリから繰返し再生ができる（▲４▼）。人の模倣音声ＢをＩＣメモリに記録すること（▲３▼）より、ＩＣメモリから人の声も繰返し再生できる（▲５▼）。音声Ａに続きＢと連続再生を繰り返すことにより、Ａ・Ｂの比較ができる。この事は人に音声の模倣学習手段を提供できる。また、ＩＣメモリに記録するため、人の学習音声を音源メディアへの録音動作は生じず、音源サンプルに書込むなどの影響はない。Reconfirm your own voice with your own ears, and realize the comparison function with the voice of the sample used as a textbook. As shown in FIG. 1, by recording a sound A from a learning sample sound source such as a tape, CD, MD, or MP3 in an IC memory (1), it can be repeatedly reproduced from the memory (4). By recording the imitation voice B of the person in the IC memory (3), the voice of the person can be repeatedly reproduced from the IC memory (5). By repeating the continuous reproduction with B following the sound A, A and B can be compared. This can provide humans with a means of learning imitation of speech. In addition, since the sound is recorded in the IC memory, the operation of recording the learning voice of the person on the sound source medium does not occur, and there is no influence of writing the sound to the sound source sample.

ＩＣメモリ中の音声データの呼出し速度を変えれば、音声の再生速度を変えることになる。音声学習トレーニングに効果的な再生速度の変更範囲を設定できる。
・速度を遅くする場合、約平常の６６％−−８３％。音声を細かく聞き取れ、分析できる。
・速度を速くする場合、約平常の１２０％−−１５０％。ヒアリングの強化訓練ができる。
ＩＣメモリに記録された音声データを上記の速度範囲で再生する場合、人の音声は歪みを最小限に抑えながら、きれいに聞き取れる。Changing the calling speed of the voice data in the IC memory changes the voice playback speed. It is possible to set a change range of the reproduction speed effective for the voice learning training.
When the speed is reduced, it is about 66% -83% of normal. Can hear and analyze voices in detail.
-When increasing the speed, about 120%--150% of normal. Hearing training can be conducted.
When audio data recorded in an IC memory is reproduced in the above-mentioned speed range, human voice can be heard clearly while minimizing distortion.

音声学習プログラム：（図３を参照）
教本となるサンプル音声（Ａ）は再生時、ＩＣメモリに自動的記録される。サンプル音声（Ａ）に一定時間の空白があれば、区切りと認識し、ＩＣメモリ中に記録されたサウンドデータＳＤ（Ａ）が自動的に数回再生。もしＳＤ（Ｂ）に人の声も吹き込んだ場合、ＳＤ（Ａ）に続きとＳＤ（Ｂ）と連続再生を数回繰り返す、ＳＤ（Ａ）とＳＤ（Ｂ）の比較になる。その後、サウンドデータをリセットし、サンプル音声（Ａ）の再生を継続。Voice learning program: (See Fig. 3)
The sample voice (A) serving as a textbook is automatically recorded in the IC memory during reproduction. If there is a blank space in the sample sound (A) for a certain period of time, it is recognized as a break and the sound data SD (A) recorded in the IC memory is automatically reproduced several times. If a human voice is also blown into SD (B), SD (A) and SD (B) will be compared with SD (A) and SD (B) will be repeated several times. After that, the sound data is reset, and the reproduction of the sample sound (A) is continued.