• Proceedings of the Acoustical Society of Korea Conference
• /
• 1998.08a
• /
• pp.298-301
• /
• 1998

HMM 훈련 방법에 따른 음소인식률의 변화에 대하여 기술한다. 음성모델은 이산 확률 밀도 혹은 연속 확률 밀도를 갖는 HMM을 사용하였으며, 훈련 알고리즘으로서는 forward-backward 와 segmental K-means 알고리즘을 사용하였다. 연속 확률 밀도는 N개의 mixture로 구성되어 있는데 1개의 mixture로 확장할 경우에서는 이진 트리 방식과 one-by-one 방식을 사용하였다. 여러 가지의 조합을 이용하여 음소인식 실험을 수행한 결과 연속 확률 분포를 사용하고 one-by-one 방식을 사용한 forward-backward 알고리즘이 가장 우수한 결과를 나타내었다.

• 한국전산유체공학회:학술대회논문집
• /
• 2011.05a
• /
• pp.14-19
• /
• 2011

Two difference cavitation models based on the homogeneous mixture model are used to study cavitating flows through converging-diverging channel. Here, the cloud cavities, pressure distributions and other results have been obtained and compared to evaluate two cavitation models. What's more, differences are observed in the simulated results, due to the differences in characteristics of each model. Analytical results shows that the new improvement cavitation model is validated to have better effects on simulating cavitating flows

• Journal of Korea Society of Industrial Information Systems
• /
• v.7 no.3
• /
• pp.9-14
• /
• 2002

In this paper we propose a method for performance enhancement of speech recognizer under noisy conditions. The parallel combination model which is presented at the PMC method using multiple Gaussian-distributed mixtures have been adapted to the variation of each mixture. The CDHMM(continuous observation density HMM) which has multiple Gaussian distributed mixtures are combined by the proposed PMC method. Also, the EM(expectation maximization) algorithm is used for adapting the model mean parameter in order to reduce the variation of the mixture density. The result of simulation, the proposed PMC adaptation method show better performance than the conventional PMC method.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2002.11b
• /
• pp.762-766
• /
• 2002

This paper describes the emotion recognition algorithm using HMM(Hidden Markov Model) method. The relation between the mechanic system and the human has just been unilateral so far. This is the why people don't want to get familiar with multi-service robots of today. If the function of the emotion recognition is granted to the robot system, the concept of the mechanic part will be changed a lot. Pitch and Energy extracted from the human speech are good and important factors to classify the each emotion (neutral, happy, sad and angry etc.), which are called prosodic features. HMM is the powerful and effective theory among several methods to construct the statistical model with characteristic vector which is made up with the mixture of prosodic features

• Annual Conference on Human and Language Technology
• /
• 2018.10a
• /
• pp.550-551
• /
• 2018

바이오마커는 체내의 상태 및 변화를 파악할 수 있는 지표이다. 이는 암을 비롯한 다양한 질병에 대하여 진단하는데 활용도가 높은 것으로 알려져 있으나, 새로운 바이오마커를 찾아내기 위한 임상 실험은 많은 시간과 비용을 소비되며, 모든 바이오마커가 실제 질병을 진단하는데 유용하게 사용되는 것은 아니다. 따라서 본 연구에서는 자연어처리 기술을 활용해 바이오마커를 발굴할 때 요구되는 많은 시간과 비용을 줄이고자 한다. 이 때 다양한 의미를 가진 어휘들이 해당 질병과 연관성이 높은 것으로 나타나며, 이들을 분류하는 것은 매우 어렵다. 따라서 우리는 Word2Vec과 가우시안 혼합 모델을 사용하여 바이오마커를 분류하고자 한다. 실험 결과, 대다수의 바이오마커 어휘들이 하나의 군집에 나타나는 것을 확인할 수 있었다.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.20 no.12
• /
• pp.3926-3937
• /
• 1996

In recent years there has been increased interest in the continuum model associated with the solidification of binary mixtures. A review of the literature, however, shows that the model verification was not sufficient or only qualitative. Present work is conducted for the reassessment of continuum model on the solidification problems of binary mixtures widely used for model validation. In spite of using the same continuum model, the results do not agree well with those of Incropera and co-workers which are benchmark problems typically used for validation of binary mixture solidification. Inferring from the agreement of present results with the analytic, experimental and other model's numerical results, this discrepancy seems to be caused by numerical errors in applying continuum model developed by Incropera and co-workers, not by the model itself. Careful examination should be preceded before selecting validation problems.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2012.11a
• /
• pp.148-151
• /
• 2012

범용 GPU (general-purpose computing on graphics processing units, GPGPU)는 GPU를 일반적인 목적으로 사용하고자 하는 병렬 컴퓨터 구조로써, 과학 연산 등 여러 분야에서 응용 프로그램의 성능을 향상시키기 위하여 사용되고 있다. 본 연구에서는 음성인식기에서 주로 사용되는 가우시안 혼합 모델(Gaussian mixture model, GMM)에서 많은 연산시간을 차지하는 관측확률 계산의 성능을 향상시키고자 GPGPU를 이용하는 알고리즘을 구현하였으며, 기존 CPU 기반 알고리즘 대비 약 13배 연산시간을 단축하였다.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.41 no.4
• /
• pp.31-37
• /
• 2004

Wavelet transform is a useful tool for analysis and process of image. This showed good performance in image compression and noise reduction. Wavelet coefficients can be effectively modeled by hidden Markov tree(HMT) model. However, in application of HMT model to image interpolation, training procedure is needed. Moreover, the parameters obtained from training procedure do not match input image well. In this paper, the structure of HMT is used for image interpolation, and the parameters of HMT are obtained from statistical characteristics across wavelet subbands without training procedure. In the proposed method, wavelet coefficient is modeled as Gaussian mixture model(GMM). In GMM, state transition probabilities are determined from statistical transition characteristic of coefficient across subbands, and the variance of each state is estimated using the property of exponential decay of wavelet coefficient. In simulation, the proposed method shows improvement of performance compared with conventional bicubic method and the method using HMT model with training.

• Korean Chemical Engineering Research
• /
• v.43 no.3
• /
• pp.371-379
• /
• 2005

Adsorption equilibria of the gases $H_2$, $CH_4$, and $C_2H_4$ and their binary mixtures on activated carbon (Calgon co.) have been measured by static volumetric method in the pressure range of 0 to 18 atm at temperatures of 293.15, 303.15, and 313.15 K. From the parameters obtained from single component adsorption isotherm, multi-component adsorption equilibria could be predicted and compared with experimental data. The binary experimental data were applied to four models : extended Langmuir, extended Langmuir-Freundlich, Ideal Adsorbed Solution theory (IAST), and Vacancy Solution Model (VSM). The models were found to describe the experimental data with a reasonable accuracy. Extended L-F model predicts equilibria of mixture better than any other model.

• Journal of Physiology & Pathology in Korean Medicine
• /
• v.19 no.1
• /
• pp.159-166
• /
• 2005

We observed the efficacy of natural herbs and mixture in treating atopic dermatitis using anti-human IgE treated Human HMC-I cell and NCNga mice model. First, we selected three herbs, Cynonchum witfordii, Diospyros kaki, Ilite which were used to treat skin disease in Traditional Korea Medicine. Using Human HMC-I cell treated with anti-human IgE, we investigate in vitro whether each herb effects on IL-4, IL-13, $TNF-{\alpha}$ expression and $TNF-{\alpha}$, Histamine secretion value. Finally, we conducted study whether the mixture of the selected herbs is more effective than each herb which consist the mixture and control group. The results show that the mixture is better in improving atopic dermatitis condition.