• Bulletin of the Korean Chemical Society
• /
• v.14 no.1
• /
• pp.95-100
• /
• 1993

A noise-induced transition is presented for a bistable system subjected to a multiplicative random force, which is singular at the unstable state. The stationary probability distribution is obtained from the Fokker-Planck equation and the effects of the singularity is analyzed. On the basis of noise-induced phase transition with Gaussian white noise, the relaxation time and the transition rate of the system are evaluated up to the first order correction of D. In the parameter region v < l, the transition rates decrease as the exponent v goes to 1 and as the coefficient of the linear term of the kinetic equation increases.

• Bulletin of the Korean Chemical Society
• /
• v.16 no.11
• /
• pp.1113-1118
• /
• 1995

The Schlo'gl model with the first order transition for a photochemical reaction is considered to study the dynamic behaviors in the neighborhood of the Gaussian white noise by obtaining the explicit results of the time-dependent variance and time correlation function with the aid of approximate methods based on the stationary properties of the system. Then, we discuss the effect of external noise strength on the stability of the model at steady states in detail.

• Bulletin of the Korean Chemical Society
• /
• v.16 no.11
• /
• pp.1119-1121
• /
• 1995

The method for the Schlo"gl model with the first order transition is extended to the Scho;gl model with the second order transition for a photochemical reaction. We obtain the explicit results of the time-dependent average and the time correlation function at the unstable steady state of the model in the neighborhood of the Gaussian white noise and then discuss the effect of noise on the dynamic properties.

• Bulletin of the Korean Mathematical Society
• /
• v.56 no.3
• /
• pp.631-648
• /
• 2019

In this paper, we consider a semilinear stochastic heat equation driven by an additive fractional white noise. Under the pathwise uniqueness property, we establish various strong stability results. As a consequence, we give an application to the convergence of the Picard successive approximation.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.23 no.5
• /
• pp.726-731
• /
• 1986

An effective filtering algorithm which can reduce noise and preserve edges for the restoration of an image degraded by additive white Gaussian noise is presented. The algorithm proposed in this paper is an extension of Lee's algorithm modified to use local gradient information as well as local statistics. It does not require image modeling, and removes noise along the orientaiton of edges so that it does not blur the edge.

• Journal of Electrical Engineering and Technology
• /
• v.11 no.6
• /
• pp.1793-1801
• /
• 2016

Speech enhancement is a challenging problem due to the diversity of noise sources and their effects in different applications. The goal of speech enhancement is to improve the quality and intelligibility of speech by reducing noise. Many research works in speech enhancement have been accomplished in English and other European Languages. There has been limited or no such works or efforts in the past in the context of Tamil speech enhancement in the literature. The aim of the proposed method is to reduce the background noise present in the Tamil speech signal by using wavelets. New modified thresholding function is introduced. The proposed method is evaluated on several speakers and under various noise conditions including White Gaussian noise, Babble noise and Car noise. The Signal to Noise Ratio (SNR), Mean Square Error (MSE) and Mean Opinion Score (MOS) results show that the proposed thresholding function improves the speech enhancement compared to the conventional hard and soft thresholding methods.

• ETRI Journal
• /
• v.20 no.2
• /
• pp.214-230
• /
• 1998

We investigate noise-induced transitions in active rotator model with a fluctuating threshold in the presence of an additive noise. The fluctuation of the threshold depends on the additive noise in a nonlinear fashion. In the white-noise limit of the fluctuation, the Fokker-Planck equation of the system reduces to that of the system with correlated linear fluctuation implying that the nonlinearity may be transformed into the correlation of linear noises. We also investigate the system with a nonlinear colored noise which depends on the additive noise as its square. The system shows a single peak, two peaks, and three peaks in its steady state probability distribution according to the noise intensities and the correlation time whose change leads to peak-creating, peak-splitting, and peak-merging transitions.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2007.10a
• /
• pp.367-370
• /
• 2007

Both impluse noise and AWGN (additive white Gaussian noise) are easily corrupted into images, during signal transmission and acquisition. Thus, an algorithm for removing both noises is represented in this paper. An impulse noise detection step can effectively separate impulse noise with AWGN, then in the noise filtering step, by using several parameters, not only impulse noise but also AWGN can be reduced. The value of those parameters are automatically changeable when the standard deviation of AWGN, the impulse noise density, and the spatial distances between pixels are different. Results of computer simulations show that the proposed approach performs better than other conventional filters.

• The Transactions of the Korean Institute of Electrical Engineers P
• /
• v.51 no.3
• /
• pp.120-125
• /
• 2002

The switching-mode power converter has been widely used because of its features of high efficiency and small weight and size. These features are brought by the ON-OFF operation of semiconductor switching devices. However, this switching operation causes the surge and EMI(Electromagnetic Interference) which deteriorate the reliability of the converter themselves and entire electronic systems. This problem on the surge and noise is one of the most serious difficulties in AC-to-DC converter. In the switching-mode power converter, the output voltage is generally controlled by varying the duty ratio of main switch. When a converter operates in steady state, duty ratio of the converter is kept constant. So the power of switching noise is concentrated in specific frequencies. Generally, to reduce the EMI and improve the immunity of converter system, the switching frequency of converter needs to be properly modulated during a rectified line period instead of being kept constant. Random Pulse Width Modulation (RPWM) is performed by adding a random perturbation to switching instant while output-voltage regulation of converter is performed. RPWM method for reducing conducted EMI in single switch three phase discontinuous conduction mode boost converter is presented. The more white noise is injected, the more conducted EMI is reduced. But output-voltage is not sufficiently regulated. This is the reason why carrier frequency selection topology is proposed. In the case of carrier frequency selection, output-voltage of steady state and transient state is fully regulated. A RPWM control method was proposed in order to smooth the switching noise spectrum and reduce it's level. Experimental results are verified by converter operating at 300V/1kW with 5%~30% white noise input. Spectrum analysis is performed on the Phase current and the CM noise voltage. The former is measured with Current Probe and the latter is achieved with LISN, which are connected to the spectrum analyzer respectively.

• Convergence Security Journal
• /
• v.24 no.2
• /
• pp.201-207
• /
• 2024

Voice is a critical element of human communication, and the development of speech recognition models is one of the significant achievements in artificial intelligence, which has recently been applied in various aspects of human life. The application of speech recognition models in the military field is also inevitable. However, before artificial intelligence models can be applied in the military, it is necessary to research their vulnerabilities. In this study, we evaluates the military applicability of the multilingual speech recognition model "Whisper" by examining its vulnerabilities to battlefield noise, white noise, and adversarial attacks. In experiments involving battlefield noise, Whisper showed significant performance degradation with an average Character Error Rate (CER) of 72.4%, indicating difficulties in military applications. In experiments with white noise, Whisper was robust to low-intensity noise but showed performance degradation under high-intensity noise. Adversarial attack experiments revealed vulnerabilities at specific epsilon values. Therefore, the Whisper model requires improvements through fine-tuning, adversarial training, and other methods.