• Proceedings of the IEEK Conference
• /
• 1998.10a
• /
• pp.1273-1276
• /
• 1998

The HMM-Net is an architecture for a neural network that implements a hidden Markov model (HMM). The architecture is developed for the purpose of combining the discriminant power of neural networks with the time-domain modeling capability of HMMs. Criteria used for learning HMM-Net classifiers are maximum likelihood (ML), maximum mutual information (MMI), and minimization of mean squared error(MMSE). In this paper we propose an efficient learning method of HMM-Net classifiers using hybrid criteria, ML/MMSE and MMI/MMSE, and report the results of an experimental study comparing the performance of HMM-Net classifiers trained by the gradient descent algorithm with the above criteria. Experimental results for the isolated numeric digits from /0/ to /9/ show that the performance of the proposed method is better than the others in the respects of learning and recognition rates.

• The Transactions of the Korean Institute of Electrical Engineers P
• /
• v.64 no.4
• /
• pp.291-296
• /
• 2015

Transmission line is exposed to a large area, and then faults are likely to occur than the other component of power system. When a fault occurs on a transmission line, fault locator helps fast recovery of power supply on power system. This paper deals with the design of a digital fault locator for improvement accuracy of the fault distance estimation and a fault occurrence position for transmission line. The algorithm of a fault locator uses a DC offset removal filter and DFT filter. The algorithm utilizes a fault data of GPS time synchronized. The computed fault information is transmitted to the other side substation through communication. The digital fault locator includes MPU module, ADPU module, SIU module, and a power module. The MMI firmware and software of the fault locator was implemented.

• ETRI Journal
• /
• v.26 no.4
• /
• pp.344-350
• /
• 2004

We propose a filter-free wavelength conversion using a Mach-Zehnder interferometer with monolithically integrated $2{\times}2$ multimode interference semiconductor optical amplifiers (MMI-SOAs). The device has been optimized by considering a non-homogeneous carrier distribution due to the self-imaging properties of the MMI-SOA. Static measurements show an extinction ratio of up to 18 dB and an input signal rejection ratio of up to 20 dB.

• Journal of the Korean Institute of Telematics and Electronics C
• /
• v.34C no.9
• /
• pp.61-67
• /
• 1997

The HMM-Net is an architecture for a neural network that implements a hidden markov model(HMM). The architecture is developed for the purpose of combining the classification power of neural networks with the time-domain modeling capability of HMMs. Criteria which are used for learning HMM_Net classifiers are maximum likelihood(ML), maximum mutual information (MMI), and minimization of mean squared error(MMSE). In this classifiers trained by the gradient descent algorithm with the above criteria. Experimental results for the isolated numbers from /young/to/koo/ show that in the binary inputs the performance of MMSE is better than the others, while in the fuzzy inputs the performance of MMI is better than the others.

• Proceedings of the KSPS conference
• /
• 2005.04a
• /
• pp.53-56
• /
• 2005

In this paper, we propose a transformation based robust adaptation technique that uses the maximum mutual information(MMI) estimation for the objective function and the linear spectral transformation(LST) for adaptation. LST is an adaptation method that deals with environmental noises in the linear spectral domain, so that a small number of parameters can be used for fast adaptation. The proposed technique is called MMI-LST, and evaluated on TIMIT and FFMTIMIT corpora to show that it is advantageous when only a small amount of adaptation speech is used.

• 제어로봇시스템학회:학술대회논문집
• /
• 1990.10b
• /
• pp.1430-1435
• /
• 1990

Using eye mark recorder and polygraph, basic laboratory experiments have been conducted to investigate the validity of psycho-physiological measures to the analysis of on-line cognitive information characteristics at man-machine interface (MMI). It is concluded that various psycho-physiological measures are useful to estimate various aspects of human cognitive characteristics and thus to the evaluation of MMI designing from the viewpoint of its harmony to the human cognitives.

• Proceedings of the Korean Nuclear Society Conference
• /
• 1995.10a
• /
• pp.117-122
• /
• 1995

An Integrated Test Facility(ITF) is a human factors experimental environment to evaluate an advanced man machine interface(MMI) design. The ITF includes a human machine simulator(HMS) comprised of a nuclear power plant function simulator, man-machine interface, experiment control station for the experiment control and design, human behavioural data measurement system, and data analysis and experiment evaluation supporting system(DAEXESS). The most important features of ITF is to secure the flexibility and expandibility of Man Machine Interlace(MMI) design to change easily the environment of experiments to accomplish the experiment's objects In this paper, we describe a development scope and characteristics of the ITF such as, hardware and software development scope and characteristics, system thermohydraulic modelling characteristics, and experiment station characteristics for the experiment variables design and control, to be used as an experiment environment for the evaluation of VDU-based control room.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2007.10a
• /
• pp.373-376
• /
• 2007

This paper presents dynamic (on-line) signature verification using direction and speed information as feature information. We describe a detail method of the adjusted constant(W) in the suggested DTW for the speed between adjacent two points.

• Proceedings of the KIEE Conference
• /
• 1988.07a
• /
• pp.866-869
• /
• 1988

In the operation of a power system, the security of the system has acquired significant importance to supply electric power of better quality. The State Estimator, a part of security functions, provides a complete real time solution estimate of the steady-state conditions of the power system for use by the Real Time Network Analysis functions. This paper briefly introduces the Fast Decoupled Weighted Least Square State Estimator which is adopted in the KEPCO EMS with features of Man-Machine Interface.

• Journal of the Earthquake Engineering Society of Korea
• /
• v.20 no.7_spc
• /
• pp.517-526
• /
• 2016

In this paper, some popular intensity measures of earthquakes including magnitude, MMI, and PGA as well as their empirical relationships are briefly reviewed since they have been widely used without prudence by mass media, the public, and even the government when asking or expressing the seismic capacity of buildings. The basic concept of current seismic design is also presented in order to facilitate relevant discussions. It is emphasized that expressing the building seismic capacity simplistically in terms of seismological quantities or terminologies like magnitude and MMI is inherently irrational, may be misleading the stakeholders, and should be avoided. Alternative expressions, more rational and consistent with current seismic design philosophy and practice, are recommended.