• 음성과학
• /
• 제10권2호
• /
• pp.35-43
• /
• 2003

General speaker verification systems improve their recognition performances by normalizing log likelihood ratio, using a speaker model and its background speaker model that are required to be verified. So these systems rely heavily on the availability of much speaker independent databases for background speaker model design. This constraint, however, may be a burden in practical and portable devices such as palm-top computers or wireless handsets which place a premium on computations and memory. In this paper, new approach for the GMM-based background model design used in portable speaker verification system is presented when the enrollment data is available. This approach is to modify three parameters of GMM speaker model such as mixture weights, means and covariances along with reduced mixture order. According to the experiment on a 20 speaker population from YOHO database, we found that this method had a promise of effective use in a portable speaker verification system.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송공학회 2009년도 IWAIT
• /
• pp.450-455
• /
• 2009

Amplitude-only logarithmic Radon transform (ALR transform) for pattern matching is proposed. This method provides robustness for object translation, scaling, and rotation. An ALR image is invariant even if objects are translated in a picture. For the object scaling and rotation, the ALR image is merely translated. The objects are identified using a phase-only matched filter to the ALR image. The ratio of size, the difference of rotation angle, and the position between the two objects are detected. Our pattern matching procedure is described, herein, and its simulation is executed. We compare matching scores with the Fourier-Mellin transform, and the general phase-only matched filter.

• Journal of Pharmaceutical Investigation
• /
• 제5권4호
• /
• pp.17-23
• /
• 1975

복용량이 비교적 적고, 난용성 의약품(醫藥品)으로 antirheumatism에 사용되고 있는 phenylbutazone을 macromolecule polymer로서 water soluble carrier인 polyvinylpyrrolidone과 solvent method로 1:1, 1:5, 및 1:9(w/w)의 coprecipitate를 형성(形成)시켰으며, 이들 coprecipitate의 용출 속도를 Pure drug 및 coprecipitate 형성 용매인methanol에서 재결정한 recrystallized pure drug의 그것과 측정 비교(比較)하였다. 1:1,1:5 및 1:9(w/w)의 coprecipitate는 recrystallized pure phenylbutazone보다 약 4.5배의 용출의 증가를 보였고, 이들 1:1,1:5,1:9(w/w)에서의 그 carrier의 양(量)에 따른 용출에의 영향은 거의 없었다. 시간(時間)에 대(對)한 log probit를 plot하여 구(求)한 dissolution half life, $T_{50%}$는 coprecipitate ratio 1:1(w/w)에서는 5.5분, 1:5에서는 10분, 1:9에서는 12.5분이었다.

• 천문학회지
• /
• 제39권4호
• /
• pp.125-128
• /
• 2006

In order to determine the precise effective temperature and surface gravity of warm stars, all synthetic spectral lines in the wavelength range of $4000-5700{\AA}$ with T=6000-7750 K, and log g=3.5, 4.0, and 4.5 for [M/H]=0.0, $V_{rot}$=10 km $s^{-1}$, and $V_{tubl}$=2 km $s^{-1}$ were calculated using the SYNSPEC package(Hubeny, et al., 1995) and the Kurucz(1995) model. Then, the depth-ratios for all line pairs were investigated and we selected two and six depth-ratios appropriate for the surface gravity and temperature indicators, respectively. We plotted six grids with X- and Y-axes for the depth-ratios of surface gravity and temperature, respectively, for the simultaneous estimation of these two atmospheric parameters. This method was applied to the spectum of $\delta$ Scu for the determination of its temperature and surface gravity simultaneously.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2002년도 하계종합학술대회 논문집(1)
• /
• pp.323-326
• /
• 2002

In this paper, we study the improved bounds on the performance of low-density parity-check (LDPC) codes over binary-input additive white Gaussian noise (AWGN) channels with belief propagation (BP) decoding in log domain. We define an extended Gallager ensemble based on a new method of constructing parity check matrix and make use of this way to improve upper bound of LDPC codes. At the same time, many simulation results are presented in this paper. These results indicate the extended Gallager ensembles based on Hamming codes have typical minimum distance ratio, which is very close to the asymptotic Gilbert Varshamov bound and the superior performance which is better than the original Gallager ensembles.

• Communications for Statistical Applications and Methods
• /
• 제22권4호
• /
• pp.313-320
• /
• 2015

In this research, we compare the efficiency of two test procedures proposed by Prentice and Gloeckler (1978) and Park and Hong (2009) for grouped data with possible right censored observations. Both test statistics were derived using the likelihood ratio principle, but under different semi-parametric models. We review the two statistics with asymptotic normality and consider obtaining empirical powers through a simulation study. The simulation study considers two types of models the location translation model and the scale model. We discuss some interesting features related to the grouped data and obtain null distribution functions with a re-sampling method. Finally we indicate topics for future research.

• 자원ㆍ환경경제연구
• /
• 제19권1호
• /
• pp.101-127
• /
• 2010

자연자원인 갯벌에 비시장가치평가법 중의 하나인 조건부가치평가법을 적용하여 근접효과를 검증하고자 하였다. 전국의 가구 수에 비례하여 1,000가구를 임의추출하였다. 그리고 갯벌로부터의 거주거리에 따라 30분 간격으로 I 구역에서 V구역까지 구역을 나누어서 표본을 재구성하였다. 이러한 근접효과 분석에는 모형의 적합성과 변수의 통계적 유의성을 고려하여 로그선형모형(log-linear)을 적용하였다. 일반적으로는 로짓모형이 많이 사용되지만 소득효과를 제어할 수 없기 때문에, 소득효과를 제어할 수 있는 로그선형 모형을 이용하였다. 결과를 살펴보면, 각 구간별 지불의사액의 차이를 명확하게 보여주었으며(${\theta}_1$ >0. ${\theta}_2$ >0), 또한 모형에서 더미변수들이 통계적으로 5% 유의수준에서 유의하다는 결론($dum1{\neq}0$, $dum2{\neq}0$, $dum3{\neq}0$, $dum4{\neq}0$)을 얻을 수 있었다. 실제로 개방형 설문조사에서도 I 구역 거주자들의 지불의사액이 V구역 거주자들의 지불의사액보다 약 26% 이상 크게 나타났다. 따라서 이렇게 도출된 결과로부터 비시장재화의 특징을 지닌 자연자원에 대하여 비시장가치평가법을 적용하는 경우에는 근접효과가 존재하기 때문에 설문에 이를 충분히 반영(embedded)하여야 한다는 결론을 도출하였다. 이를 통해서 좀 더 명확한 지불의사액 추정이 가능하며, 편익이전시 과소평가되는 문제를 예방할 수 있다.

• 재활복지공학회논문지
• /
• 제3권1호
• /
• pp.27-34
• /
• 2009

난청인의 청력 손실은 개인별로 다양하므로 기존의 대칭적으로 주파수 밴드별 청각 필터를 구현하는 방법은 다양한 형태의 난청인의 청력 손실을 적절하게 모사해주지 못한다. 각 중심주파수와 음성의 입력크기에 따라 청각 필터의 형태가 비대칭적으로 바뀌고 청력손실이 있는 난청인은 청력 손실에 따라 청각필터의 형태가 정상인들과는 다른 형태로 바뀌게 되며 음질에도 차이가 있다. 본 연구에서는 난청인의 난청 정도에 따라 변하는 비대칭 청각 특성을 잘 반영한 청각필터를 구현하여 몇 가지 실험을 통해 각 구현된 청각 필터의 성능을 객관적으로 평가하였다. 실험은 구현된 청각 필터를 통한 음성의 perceptual evaluation of speech quality (PESQ) 와 log likelihood ratio (LLR)를 사용하였으며 그 값을 통해 처리된 음성의 객관적인 음질과 왜곡정도를 평가 하였다. 청력 손실을 주었을 때 대칭과 비대칭 청각 필터사이의 PESQ 와 LLR 값을 실험해 본 결과 청각 필터 간의 큰 차이를 보였다. 위 실험 결과들로 대칭과 비대칭 청각 필터의 형태에 따라서 음성의 음질에 영향을 받는다는 것을 알 수 있었다. 특히, 난청이 있을 때 중심 주파수별 청각 필터의 비대칭적 형태 변화가 난청인이 받아들이는 음질에 영향이 있었다.

• 수산해양기술연구
• /
• 제50권4호
• /
• pp.427-434
• /
• 2014

Although ex situ target strength (TS) measurements using dual- and split-beam systems have become the primary approach of estimating fish abundance, theoretical model estimation is a powerful tool for verifying the measurements, as well as for providing values when making direct measurements is difficult. TS values for 20 samples of live bambooleaf wrasse (Pseudolabrus japonicus) whose target length (TL) ranged between 13.7 and 21.3 cm were estimated theoretically using the Kirchhoff-ray mode model, and the TS values for 18 live fish samples were additionally measured at ${\sim}0^{\circ}$ tilt angle to the swimming aspect using a tethered method at a frequency of 120 kHz to verify the theoretical values. The digitizing intervals used to extract the fish body and swim bladder morphology in the X-ray photographs significantly affected the calculated TS patterns, but variations based on the speed of sound and density ratio values for the general range of fish flesh were relatively small (within 1 dB). Close agreement was observed between the measured and theoretical TS values, and the correlation between the average TS and body length of the fish could be calculated accurately as <$TS_{120kHz}$>= 20logTL (cm) -71.6 using the theoretical method.

• Journal of Nutrition and Health
• /
• 제30권3호
• /
• pp.349-359
• /
• 1997

This study was designed to study sources of variability(subjects, day by day, season, day of week, sequence, number of repeated days, compliance etc.) in multiple 24-hour dietary recall method. Dietary intakes of college women in Seoul were obtained using 24-hour dietary recalls, consecutive 7 days, 4 times for 1 year. The result of this study shows that 1) the reproducibility of multiple 24-hour recalls is low(intraclass correlation coefficient<0.3), 2) the intraindividual variations of all nutrients are more than 80%, 3) the major components of variation are interindividual vriability and intraindividual variability(within person variability and methodologic error). There were small but statistically significant season and compliance effects. 4) the ratio within-person varation to between-person variation for absolute and transformed nutrient intakes except some nutrients such as vitamin A, vitamin C, energy from protein and fat etc, was less than 1.7 the minimum numbers of days needed in 24-hour dietary recalls to estimate usual intake for this group were 1 to 21 days. And the numbers of days needed to get the usual intake of an individual varied among individuals and within individuals for different nutrients, ranged from 1 days for log carbohydrate(% energy) to 470 days for Na. 6) There were greater than 0.7 correlation coefficients between the average nutrient intakes of 12 days and intakes of 28 days. Therefore, it was desirable to take the 24-hour recall more than 12 times, repeating 3 days every season.