• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1998.06a
• /
• pp.719-724
• /
• 1998

This paper presents a systematic developement of a formal approach to inference in approximate reasoning. We introduce some measures of similarity and discuss their properties. Using the concept of similarity index we formulate two methods for inferring from vague knowledge. In order to illustrate the effectiveness of the proposed technique we use it to develop a vowel recognition system.

• Proceedings of the KOSOMBE Conference
• /
• v.1994 no.05
• /
• pp.79-81
• /
• 1994

Petubation in the fundamental frequency and in the peak amplitude of the EGG signal derived with a four-electrode EGG system were investigated for the purpose of developing useful measures for the detection of layngeal pathology. The data were compared to the degree of amplitude perturbation and frequency perturbation. There was a close relation between amplitude perturbation and frequency perturbation analysis of EGG signal and degree of laryngeal pathology.

• Proceedings of the Korean Information Science Society Conference
• /
• 2004.10a
• /
• pp.784-786
• /
• 2004

정보통신 단말기에서 한글을 입력할 때는 단말기에 부착되어 있는 한정된 개수의 입력키만 을 이용해야 한다는 제약이 있다. 이러한 제약 조건 하에서 한글 모음을 입력하는 효율적인 방법을 모색하기 위해 기본 모음 집합의 크기를 변화시키면서 모음 조합 방법을 제시하였다. 그 결과로 8개의 기본 모음 'ㅏㅓ ㅗ ㅜ ㅡ ㅣ ㅐ ㅔ' 로부터 모음의 조합 원리 및 가획 원리에 의해 입력키가 할당되지 않은 모음을 조합하는 방식을 제안하였다. 이 방법은 유사 모음간의 전환이 가능한 오류 수정 방식과 다중 입력 방식을 도입하여 사용자 편의성을 추구하였다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2004.04b
• /
• pp.907-909
• /
• 2004

정보화 시대가 진전됨에 따라 휴대폰과 리모콘 등 휴대용 단말기에서 한글을 입력할 필요성이 증가하고 있다. 휴대용 단말기에서 한글 입력 방식은 단말기에 부착되어 있는 제한된 버튼만을 이용해야 한다는 제약이 있다. 이러한 제약 조건 하에서 한글 모음을 편리하고 빠르게 입력할 수 있는 방법으로 'ㅏ/ㅓ/ㅗ/ㅜ/ㅡ/ㅣ/ㅐ/ㅔ'를 기본 요소로 하여 조합되는 모음의 입력 방식을 다양화하였으며, 기존의 천지인, 나랏글 방식과 비교했을 때 사용자 편의성과 입력의 신속성 측면에서 매우 효율적임을 확인하였다.

• The Transactions of the Korea Information Processing Society
• /
• v.7 no.6
• /
• pp.1964-1972
• /
• 2000

For the Korean language, on acoustics, it creates a different form of phonetic value not a phoneme by its own peculiar property. Therefore, the construction of extended recognition system for understanding Korean language should be created with a study of the Korean rule-based system, before it can be used as post-processing of the Korean recognition system. In this paper, text-based Korean rule-based system featuring Korean peculiar vocal sound changing rule is constructed. and based on the text-based phonetic value result of the system constructed, a preliminary phonetic value segmentation border points with non-uniform blocks are extracted in Korean isolated word speech. Through the way of merge and recognition of the non-uniform blocks between the extracted border points, recognition possibility of Korean voice as the form of the phonetic vale has been investigated.

• Speech Sciences
• /
• v.13 no.4
• /
• pp.23-42
• /
• 2006

This paper reports a preliminary result of an acoustic study on the stress and intonational system in Lakhota, a native American language. It investigates how the stress and intonation in Lakhota are phonetically manifested; and how the stress interacts with other prosodic factors. The results preliminarily obtained from one native Lakhota speaker suggest that the primary cue of the stress is relatively high F0 which is often accompanied by higher intensity (for the vowel) and longer VOT (for aspirated stops). The results also indicate that stress is not reliably marked by duration. The stress system, however, interacts with the intonational pattern, such that, for example, intonational peak falls on the stressed syllable with a general pattern of L+H* and that it interacts with the boundary tone L%, resulting in mid tone utterance-finally. This paper can be viewed largely as a qualitative study on an understudied native American language, Lakhota and as forming a basis for further development of its stress and intonation system whose acoustic properties of its prosodic system have not been investigated before.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.14 no.2
• /
• pp.721-727
• /
• 2013

This study sought to compare the difference of Phonation Threshold Pressure and Nasalance according to vowels. We tested 28 normal females(20~30age) who be in attendance at a school from May to August 2012. We measured Phonation Threshold Pressure through PAS(Phonatory aerodynamic system, model 6600, KAY electronics, Inc) that aero-mechanical instrument called. Phonation Threshold Pressure was measured by using the consonant /p/ and the three vowels /a, i, u/ and was analyzed the three times of Phonation the five times. Also We measured Nasalance through the Nasometer${\Pi}$(model 6450, KAY electronics, Inc). The vowel's phonation was recored by using the three vowels /a, i, u/. the nasalance score was measured via Nasometer. As a Result, Phonation Threshold Pressure according to three vowels show significantly difference. Also, Nasalance according to three vowels show significantly difference. Finally, Correlation coefficient of Phonation Threshold Pressure and Nasalance according to three vowels show positive relation ship. Therefore Phonation Threshold Pressure and Nasalance according to three vowels have closely relationship.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.19 no.12
• /
• pp.2364-2375
• /
• 1994

For the efficient recognition of printed Korean characters, a multiplexed minimum average correlation energy(MMACE) filter is proposed. Proposed method solved the disadvantages of the tree structure algorithm which recognition system is very huge and recognition method is sophisticated. Using only one consonant MMACE filter and one vowel one, we recognized the full Korean character. Each MMACE filter is multiplexed by 4 K-tuple MACE filters which are synthesized by 24 consonants and vowels. Hence the proposed MMACE filter and the correlation distribution plane are divided by 4 subregion. We obtained the binary codes for the Korean character recognition from each correlation distribution subplane. And the obtained codes are compared with the truth table for consonants and vowels in computer. We can recognize the full Korean characters when substitute the corresponded consonant or vowel font of the consistent code to the correlation peak place in the output correlation plane. The computer simulation and optical experiment results show that the proposed compact Korean character recognition system using the MMACE filters has high discrimination capability.

• Speech Sciences
• /
• v.9 no.4
• /
• pp.49-58
• /
• 2002

This study was designed to examine the effects of vocal loudness on Nasalance scores in Korean vowels. One normal adult without 'hypemasality' produced 3 types of vowels (/i/, /a/, /u/) in 3 levels of loudness (60, 70, 80dB). Nasalance scores (%) and oral & nasal energy intensity (dB) were measured through NasalView system. These data were compared among 3 levels of loudness and among 3 types of vowels. The results were as follow: For all vowels, Nasalance scores significantly reduced when vocal loudness increased. Because the increases of oral energy intensity were larger than that of nasal energy intensity. For all levels of loudness, Nasalance scores significantly differed among vowel types (/i/>/a/>/u/).

• Journal of Korea Multimedia Society
• /
• v.22 no.8
• /
• pp.852-859
• /
• 2019

With the development of IT technology and smart devices, various applications utilizing image information are being developed. In order to provide an intuitive interface for pronunciation recognition, there is a growing need for research on pronunciation recognition using mouth feature values. In this paper, we propose a system to distinguish Korean vowel pronunciations by detecting feature points of lips region in images and applying Bayesian based learning model. The proposed system implements the recognition system based on Bayes' theorem, so that it is possible to improve the accuracy of speech recognition by accumulating input data regardless of whether it is speaker independent or dependent on small amount of learning data. Experimental results show that it is possible to effectively distinguish Korean vowels as a result of applying probability based Bayesian classification using only visual information such as mouth shape features.