• The Journal of Korean Association of Computer Education
• /
• v.6 no.2
• /
• pp.9-21
• /
• 2003

This paper proposes an automatic construction system for transliteration dictionary from English-Korean parallel corpus. The system works in 3 steps: it extracts all nouns from Korean documents as the first step, filters transliterated foreign word nouns out of them with the language identification method as the second step, and extracts the corresponding English words by using a probabilistic alignment method as the final step. Specially, the fact that there is a corresponding English word in most cases, is utilized to extract the purely transliterated part from a Koreans word phrase, which is usually used in combined forms with Korean endings(Eomi) or particles(Josa). Moreover, the direct phonetic comparison is done to the words in two different alphabet systems without converting them to the same alphabet system. The experiment showed that the performance was influenced by the first and the second preprocessing steps; the most efficient model among manually preprocessed ones showed 85.4% recall, 91.0% precision and the most efficient model among fully automated ones got 68.3% recall, 89.2% precision.

• Journal of the Korean Institute of Telematics and Electronics C
• /
• v.35C no.6
• /
• pp.76-87
• /
• 1998

This paper deals with a system which recognizes dynmic hand gestures, Korean Sign Language(KSL), mixed with static hand gesture, Korean Manual Alphabet(KMA), continuously. Recognition of continuous hand gestures is very difficult for lack of explicit tokens indicating beginning and ending of signs and for complexity of each gesture. In this paper, state automata is used for segmenting sequential signs into individual ones, and basic elements of KSL and KMA, which consist of 14 hand directions, 23 hand postures and 14 hand orientations are used for recognition of complex gestures under consideration of expandability. Using a pair of CyberGlove and Polhemus sensor, this system recognizes 131 Korean signs and 31 KMA's in real-time with recognition rate 94.3% for KSL excluding no recognition case and 96.7% for KMA.

• Journal of Satellite, Information and Communications
• /
• v.8 no.1
• /
• pp.8-13
• /
• 2013

It is necessary to develop the next generation FTS which is suitable for our environment and effectively operates many launch vehicles. Standard tone, Secure tone, MHA, EFTS and DSSS are studied for the next generation FTS. FTS requires a high quality of performance and reliability because of their specific mission. And few FTSs are needed but the price is very expensive. Therefore we must investigate a part of the FTS whether the part can be reused for a part of the next FTS. In this paper, we use CCDF of the transmitted signal from FTS as the method to study a possibility of reusing HPA used in the present system. The simulation results show that PEP of Standard tone is 0.21dB and Secure tone and MHA has the same PEP. CPFSK's PEP is 1.81dB and PEP of DSSS using BPSK modulation is 2.6dB.

• Journal of The Korean Society of Integrative Medicine
• /
• v.3 no.1
• /
• pp.11-21
• /
• 2015

Purpose: This study intends to investigate if visual perception training program has influence on writing intelligibility and visual perception ability of the children who have spastic cerebral palsy. Methods: The subjects were 12 children of 4-8 years old. Classifying them into 2 groups - cerebral palsy children group(6) and normal children group(6), visual perception training program was conducted to the cerebral palsy children group twice a week for 4 weeks. Korean alphabet writing test and Korean-version development test on visual perception(DTVP-2) were used for the test. Result: First, the cerebral palsy children group had significantly lower scores on visual perception than the normal children group. Second, Two groups didn't show significantly difference between before and after the visual perception training program. Third, The normal children group significantly increased at three parts without the visual perception training program. Forth, The cerebral palsy children group significantly increased at the only a part after the visual perception training program Conclusion: For the visual perception ability improvement of cerebral palsy children, it is necessary to systematic and continued intervention. And the study about the effect of the visual perception training program application according to by type characteristic of the cerebral palsy child wishes to proceed.

• Journal of KIISE:Computer Systems and Theory
• /
• v.32 no.5
• /
• pp.255-259
• /
• 2005

To search a pattern P in a text, such index data structures as suffix trees and suffix arrays are widely used in diverse applications of string processing and computational biology. It is well known that searching in suffix trees is faster than suffix ways in the aspect of time complexity, i.e., it takes O(${\mid}P{\mid}$) time to search P on a constant-size alphabet in a suffix tree while it takes O(${\mid}P{\mid}+logn$) time in a suffix way where n is the length of the text. In this paper we present a linear-tim8 search algorithm in suffix arrays for constant-size alphabets. For a gene.al alphabet $\Sigma$, it takes O(${\mid}P{\mid}log{\mid}{\Sigma}{\mid}$) time.

• Journal of the Korean Society for Library and Information Science
• /
• v.12
• /
• pp.7-37
• /
• 1985

In most of the modern libraries throughout the world, it is popular to adopt the author number which arranges the books or entries in alphabetical order of their author's names within the same ultimate class. However, viewed historically, that was the Western practice never used in the East before. Traditionally chronological order was followed in the East. Book number not only individualizes the books within the same class but also makes it easy for the user to choose and find out the relevant materials. In this respect, chronological book number is decidedly superior to all kinds of book number systems ever have been existed. Especially in these days of rapid obsolescence of documents and with serious problems of storage of documents, the chronological order seems to be the most modern and future-oriented of all other book number systems because it distinguishes clearly new materials from aged materials by the date of publication and controls the stack spaces mechanically and effectively. This writer devised New Chronological Book Numbers adaptable to both the Eastern and Western materials, and the system has been adopted at Yonsei University Library. The features of the system are as follows: (1) It is easy to type the call number in cataloging the Eastern and Western materials. That is because the structure of the chronological number is pure numerical notation by taking the last two figures of the year within twentieth century(e.g. 85 for AD 1985), that is 1900s, and the last three figures of the year from twenty-first century to thirtieth century, that is 2000s. (2) It does not make all classes have chronological number uniformly. Within the classes for individual biographies, genealogies and collections of biography of specific family, books about individual organization, individual literary and artistic works, philosophical works of the famous philosophers, books and its facsmile editions published before 1900, new editions, translations, commentaries, serials, chronological number is assigned to the author number as in the past. (3) It separates the Eastern materials from the Western materials and makes the typing easy in cataloging. That is because the subarrangement within the same chronological book number is made by the accession order which uses 'panjol-ponmun' or the characters of basic syllabic table of Korean alphabet in case of the Eastern materials and Roman alphabet in case of the Western materials.

• KIPS Transactions on Software and Data Engineering
• /
• v.6 no.5
• /
• pp.267-278
• /
• 2017

As the same letter can be pronounced differently depending on word contexts, one should refer to a lexicon in order to pronounce a word correctly. Phonetic alphabets that lexicons adopt as well as pronunciations that lexicons describe for the same word can be different from lexicon to lexicon. In this paper, we use a sequence-to-sequence model that is widely used in deep learning research area in order to convert automatically from one pronunciation to another. The 12 seq2seq models are implemented based on pronunciation training data collected from 4 different lexicons. The exact accuracy of the models ranges from 74.5% to 89.6%. The aim of this study is the following two things. One is to comprehend a property of phonetic alphabets and pronunciations used in various lexicons. The other is to understand characteristics of seq2seq models by analyzing an error.

• Annual Conference on Human and Language Technology
• /
• 2023.10a
• /
• pp.659-663
• /
• 2023

이 논문은 한국어 음성인식 성능을 개선하고자 기존 음성인식 과정을 자모단위 음성인식 모델과 신경망 기반 자모 병합 모델 총 두 단계로 구성하였다. 한국어는 조합어 특성상 음성 인식에 필요한 음절 단위가 약 2900자에 이른다. 이는 학습 데이터셋에 자주 등장하지 않는 음절에 대해서 음성인식 성능을 저하시키고, 학습 비용을 높이는 단점이 있다. 이를 개선하고자 음절 단위의 인식이 아닌 51가지 자모 단위(ㄱ-ㅎ, ㅏ-ㅞ)의 음성인식을 수행한 후 자모 단위 인식 결과를 음절단위의 한글로 병합하는 과정을 수행할 수 있다[1]. 자모단위 인식결과는 초성, 중성, 종성을 고려하면 규칙 기반의 병합이 가능하다. 하지만 음성인식 결과에 잘못인식된 자모가 포함되어 있다면 최종 병합 결과에 오류를 생성하고 만다. 이를 해결하고자 신경망 기반의 자모 병합 모델을 제시한다. 자모 병합 모델은 분리되어 있는 자모단위의 입력을 완성된 한글 문장으로 변환하는 작업을 수행하고, 이 과정에서 음성인식 결과로 잘못인식된 자모에 대해서도 올바른 한글 문장으로 변환하는 오류 수정이 가능하다. 본 연구는 한국어 음성인식 말뭉치 KsponSpeech를 활용하여 실험을 진행하였고, 음성인식 모델로 Wav2Vec2.0 모델을 활용하였다. 기존 규칙 기반의 자모 병합 방법에 비해 제시하는 자모 병합 모델이 상대적 음절단위오류율(Character Error Rate, CER) 17.2% 와 단어단위오류율(Word Error Rate, WER) 13.1% 향상을 확인할 수 있었다.

• Journal of Korean Ophthalmic Optics Society
• /
• v.20 no.2
• /
• pp.177-186
• /
• 2015

Purpose: The Purpose of this study is to investigate if the relative legibility of optotypes affects the corrected visual acuity in visual acuity test. Methods: After measuring the relative legibility of 'Landolt ring target', 'arabic number target', 'alphabet target' by showing as a single-letter-target in 24 subjects without specific ocular diseases and ocular surgery experience, the relative legibility of 0.8, 1.0, 1.25 row of vision according to type of target in 7 types of chart were compared. After then we compared by measuring the corrected visual acuity according to type of target by using binocular MPMVA test (#7A) in 60 myopic subjects. Results: In 3 types of target the worst relative legibility target was 'Landolt ring target' with legible distance of $98.97{\pm}4.57cm$ and the best relative legibility target was 'alphabet target' with legible distance of $108.42{\pm}3.46cm$. There was no difference of the relative legibility according to type of chart or visual acuity level in the row of vision if other conditions are the same. In 1.0 and 1.25 row of vision the difference of relative legibility according to type of target was shown the statistically significant difference between 'Landolt ring target' and 'alphabet target' as $-0.07{\pm}0.06$ (p=0.02) and $-0.06{\pm}0.06$ (p=0.04) respectively. In myopia the difference of corrected visual acuity according to type of target was statistically significant difference between 'Landolt ring target' and 'arabic number target' as $-0.04{\pm}0.02$ (p=0.02) and it was especially remarkable in the low myopia. Conclusions: Measuring visual acuity with different optotypes could cause the errors in best vision measurement value because there was difference of the relative legibility according to type of target even though visual acuity level is same in the row of vision.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2007.11a
• /
• pp.300-303
• /
• 2007

지문자는 청각장애인이 사용하는 수화로 표현하지 못하는 한글 문자를 알파벳으로 표시하기위한 손 제스처이다. 본 논문에서는 추출된 손 영역의 무게 중심과 퍼지 논리를 이용하여 지문자를 인식하는 알고리즘을 제안하고, 한글 문자를 표현하는 시스템을 개발한다. USB 카메라로부터 얻어진 영상에서 히스토그램을 이용하여 손의 피부색 영역을 추출하고, 영상 마스크를 이용하여 피부색이 아닌 배경 영역을 제거한다. 문턱 값을 사용하여 얻어진 이진화된 영상에서 손의 영역을 검출하고, 무게 중심을 이용하여 손 중심과 손가락 끝의 거리를 측정한다. 얻어진 거리 정보에 퍼지 기법을 적용하여 손가락의 굽힘 정도를 판단하고, 손 모양 데이터베이스에서 손가락 굽힘 정도와 가장 근사한 한글 문자를 선택한다.