• Phonetics and Speech Sciences
• /
• v.1 no.4
• /
• pp.119-125
• /
• 2009

This paper proposes fast speaker adaptation method using artificially distorted speech in telematics terminal under the car noise environment based on eigenspace-based maximum likelihood linear regression (ES-MLLR). The artificially distorted speech is built from adding the various car noise signals collected from a driving car to the speech signal collected from an idling car. Then, in every environment, the transformation matrix is estimated by ES-MLLR using the artificially distorted speech corresponding to the specific noise environment. In test mode, an online model is built by weighted sum of the environment transformation matrices depending on the driving condition. In 3k-word recognition task in the telematics terminal, we achieve a performance superior to ES-MLLR even using the adaptation data collected from the driving condition.

• Journal of the Korean Society for Library and Information Science
• /
• v.9
• /
• pp.99-118
• /
• 1982

This paper first reviews various techniques proposed for automatic indexing with special emphasis placed on statistical techniques. Frequency-based statistical techniques are categorized into the following three approaches for further investigation on the basis of index term selection criteria: term frequency approach, document frequency approach, and probabilistic approach. In the experimental part of this study, Pao's technique based on the Goffman's transition region formula and Harter's 2-Poisson distribution model with a measure of the potential effectiveness of index term were tested. Experimental document collection consists of 30 agriculture-related documents written in Korean. Pao's technique did not yield good result presumably due to the difference in word usage between Korean and English. However, Harter's model holds some promise for Korean document indexing because the evaluation result from this experiment was similar to that of the Harter's.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2003.11a
• /
• pp.473-476
• /
• 2003

영한 자동번역 시스템에서 대역어 선택 모듈은 어휘 변환을 수행한다. 일반적으로 영어 단어는 다양한 한국어 단어로 번역될 수 있는 의미적 모호성을 지니고 있으며, 고품질의 영한 자동번역 결과를 제공하기 위해서는, 해당 문맥에 가장 적합한 한국어 단어가 선택되어야 한다. 본 논문에서는 영어의 명사 어휘에 대하여, 벡터를 사용하는 2 단계 영한 대역어 선택 기법을 제안한다. 벡터를 사용하는 2 단계 대역어 선택 방식은 첫 번째 단계에서, 원문에서 사용된 영어 명사의 의미를 결정하고, 두 번째 단계에서, 해당 의미를 지니는 유사 한국어 대역어 가운데, 생성될 한국어 문맥에 맞는 적합한 한국어 대역어를 선택한다. 또한 제안하는 방법의 타당성을 검증하기 위해 현재 우리가 개발중인 Tellus-EK 영한 자동번역 시스템에 적용한 결과를 논한다.

• Journal of Home Health Care Nursing
• /
• v.10 no.2
• /
• pp.113-122
• /
• 2003

Purpose: This is a simple survey for discussion about cost analysis methodological issues in home care nursing service studies. Method: The subject of this study were articles published in Korea from 1961 to August, 2002, and searched by key word 'cost' and 'nursing' from various DB(National Assembly Library, The National Library of Korea, RICH etc). Finally, 13 articles were collected. Result: 1) The major type of cost analysis studies was a cost comparison or a simple cost study. 2) The important methodological weaknesses were as followers; (1) few studies were suggested cost analysis framework or analytic perspective, (2) it ,was not enough to describe for basis of selection of cost/effectiveness items, (3) few studies were done by sensitivity analysis. Conclusion: These above results will be used to develop a more proper cost analysis methodological framework in home care nursing services and also to contribute as a guideline for further studies.

• 제어로봇시스템학회:학술대회논문집
• /
• 2002.10a
• /
• pp.70.1-70
• /
• 2002

In the fuzzy applications and theories, it is very important to consider how to design the optimal fuzzy model from short training data, in order to construct the reasonable fuzzy model for identifying the practical process. There are several concerns to be confirmed for efficient fuzzy model design. One of concern is the optimization problem of the fuzzy model. In various applications, the genetic algorithm is widely applied to obtain optimal fuzzy model and other cases that adopt evolutionary mechanism of the nature. If we use natural selection and multiplication operation of the genetic algorithm, early convergence to local minimum can be occurred. In other word, we can find only optimum...

• Annual Conference on Human and Language Technology
• /
• 2001.10d
• /
• pp.310-316
• /
• 2001

이 논문은 '에서', '으로'와 같은 한국어의 부사격 조사들을 다국어 기계번역 시스템에서 다룰 때 올바른 역어 선택을 위한 3단계 변환 방식과 이를 위한 부사격 조사의 언어학적 모델링 방법을 제시한다. 3단계 변환 방식은 부사격 조사의 의미 모호성 해소, 의사 중간언어표상 (Quasi-Interlingua Representation)으로의 변환, 전치사 선택의 3단계로 구성되어 있다. 본 논문에서 중점적으로 다루게 될 세번째 단계, 즉 영어나 독일어에서 한국어의 부사격 조사에 대한 전치사 선택의 단계에서 올바른 대역어 선정 방법론의 핵심이 되는 부사격 조사에 대한 언어학적 모델링을 위해 Pustejovsky (1995)의 생성 어휘부 이론 (Generative Lexicon Theory)을 도입한다. 이 논문에서 제시한 방법론은 그 타당성의 수학적 검증을 위해 통합기반 기계번역 시스템인 CAT2에서 구현되었으나, 방법론 자체는 특정 시스템에 제한됨 없이 범용적으로 적용될 수 있을 것이다.

• Annual Conference on Human and Language Technology
• /
• 1996.10a
• /
• pp.173-176
• /
• 1996

한국어 생성기의 성능은 여러 가지 요소로 평가될 수 있다. 속도, 생성 문장의 복잡성 등 여러 가지 요소가 평가 대상이 될 수 있다. 그 중에서 가장 중요한 요소로 평가될 수 있는 것은 생성되는 문장이 얼마나 자연스러운 것인가 하는 것이다. 자연스러움의 정도는 정확히 측정할 수 없지만 그 중에서 어절의 순서 배치, 대응되는 정확한 어휘의 선정, 조사, 어미 등의 적절한 선정을 들 수 있다. 본 논문에서는 특정한 양상을 술어가 가질 때 주격조사의 선정에 주안점을 두었다. 기존의 생성기[l,3,7,9]에서는 대표격 조사 '가(무종성)'나 '이(유종성)'를 사용하였는데 양상을 동반할 때에는 '는(무종성)'이나 '은(유종성)'을 사용하는 것이 더 자연스럽다는 것을 보이도록 하겠다.

• Journal of the Korean Institute of Telematics and Electronics A
• /
• v.31A no.4
• /
• pp.127-135
• /
• 1994

In this paper we propose an EWLD(Enhanced Weighted Levenshtein Distance) algorithm to organize code-string pattern matching linear array processor based on the mappting to an one-dimensional array from a two-dimensional matching matrix, and design a processing element(PE) of the processor, N PEs are required instead of NS02T in the processor because of the mapping. Data input and output between PEs and all internal operations of each PE are performed in bit-serial fashion. The bit-serial operation consists of the computing of word distance (WD) by comparison and the selection of optimal code transformation path, and takes 22 clocks as a cycle. The layout of a PE is designed based on the double metal $1.5\mu$m CMOS rule. About 1,800 transistors consistute a processing element and 2 PEs are integrated on a 3mm$\times$3mm sized chip.

• Journal of Korea Multimedia Society
• /
• v.5 no.2
• /
• pp.206-214
• /
• 2002

In this paper, it is designed and verified the baseband system that performs various protocol functions of specification of the Bluetooth system. In order to verify the developed circuits, various baseband functions are tested by using the ModelSim simulator. The developed circuits operate at 4MHz main clock. Test suite includes hap selection function, generation of the sync word, error correction(1/3 rate FEC, 2/3 rate FEC), HEC generation/checking, CRC generation/checking, data whitening/dewhitening and packet trans/reception procedure. etc. As a result of the simulation, it is verified that the developed baseband system conform to the specification of the Bluetooth system.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.49 no.2
• /
• pp.61-68
• /
• 2012

This paper describes a normalized numeric image descriptor used to assess the luminance and contrast of the image. The proposed image descriptor used the each pixel data as weighted value of the probability density function (PDF) and defined by normalization in order to objective represent. The proposed image numeric descriptor can be used to the adaptive gamma process because it suggests the objective basis of the gamma value selection.