• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.24 no.6B
• /
• pp.1203-1210
• /
• 1999

Coefficients of the adaptive filter are misadjusted by primary signals which are uncorrelated with reference signals of the adaptive filter. In this paper, the normalized sign algorithm is analyzed and compared with the NLMS algorithm by the steady state performance and the transient characteristics when target signals are included in primary signals. The excess mean square error of the NLMS algorithm is proportional to the power of target signals. That of normalized sign algorithm is proportional to the square root of the target signal power. However, the convergence speed of the normalized sign algorithm is slower than that of NLMS algorithm. In this paper, it is shown that theoretical analysis of the steady state performance and the transient characteristics are well consisted with the results of computer simulation.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2018.05a
• /
• pp.335-337
• /
• 2018

음성향상기법은 음성에 포함된 잡음이나 잔향을 제거하는 기술로써 마이크로폰으로 입력된 음성신호는 잡음이나 잔향에 의해 왜곡되어지므로 음성인식, 음성통신 등의 음성신호처리 기술의 핵심 기술이다. 이전에는 음성신호와 잡음신호 사이의 통계적 정보를 이용하는 통계모델 기반의 음성향상기법이 주로 사용되었으나 통계 모델 기반의 음성향상기술은 정상 잡음 환경과는 달리 비정상 잡음 환경에서 성능이 크게 저하되는 문제점을 가지고 있었다. 최근 머신러닝 기법인 심화신경망 (DNN, deep neural network)이 도입되어 음성 향상 기법에서 우수한 성능을 내고 있다. 심화신경망을 이용한 음성 향상 기법은 다수의 은닉 층과 은닉 노드들을 통하여 잡음이 존재하는 음성 신호와 잡음이 존재하지 않는 깨끗한 음성 신호 사이의 비선형적인 관계를 잘 모델링하였다. 이러한 심화신경망 기반의 음성향상기법을 향상 시킬 수 있는 방법 중 하나인 강화학습을 적용하여 기존 심화신경망 대비 성능을 향상시켰다. 강화학습이란 대표적으로 구글의 알파고에 적용된 기술로써 특정 state에서 최고의 reward를 받기 위해 어떠한 policy를 통한 action을 취해서 다음 state로 나아갈지를 매우 많은 경우에 대해 학습을 통해 최적의 action을 선택할 수 있도록 학습하는 방법을 말한다. 본 논문에서는 composite measure를 기반으로 reward를 설계하여 기존 PESQ (Perceptual Evaluation of Speech Quality) 기반의 reward를 설계한 기술 대비 음성인식 성능을 높였다.

• Journal of Korean Society of Transportation
• /
• v.24 no.5 s.91
• /
• pp.19-33
• /
• 2006

This paper presents the results of an experimental study on the application of NTCIP protocol to Korean traffic signal control system. For this study the communication Protocol of the existing traffic signal control system was adjusted to meet NTCIP standard. Management information base for Korea real-time traffic signal control system, message library of OER, traffic control center management software supporting SNMP/SFMP Protocol, and agent softwares for local controllers were developed during the experimental study. The applicability test of the adjusted system by NTCIP standard was performed. Fifty eight Percent of communication packets were lost at 2.400bps communication speed, which made the operation impossible. The experimentations with communication speeds 4,800bps and 9,600bps did not cause problems. In conclusion, to apply the NTCIP standard to domestic real-time traffic control system, communication environments need to be upgraded to 4,800bps or higher.

• Journal of Aerospace System Engineering
• /
• v.17 no.1
• /
• pp.68-78
• /
• 2023

In this paper, improvement measures were studied for false alarm (abnormal lighting) of emergency hydraulic warning light due to RF radiation signals in aircraft. When the warning lamp of the emergency hydraulic system is lit during flight due to electromagnetic interference caused by RF radiation signals, it affects flight mission and safety which can result in efficiency degradation and maintenance of flight mission. For false alarm of emergency hydraulic warning light, root cause analysis and troubleshooting were carried out. By shielding a composite strip under fuselage of the aircraft located in the path of the RF radiation signal, the RF radiation signal flowing into the emergency hydraulic indicator was blocked. Results of applying the improvement method through ground and flight monitoring were also described.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 1998.11a
• /
• pp.420-423
• /
• 1998

최근 이동 통신이나 위성통신과 같은 많은 디지털 통신 시스템에서 시스템의 용량을 증가시키고 인접 채널간 간섭을 줄이기 위해 많은 연구가 진행되고 있으며, 그 중에서 어레이 빔 형성 기법에 대한 연구가 활발하게 진행되고 있다. 이에 본 논문에서는 원하는 신호의 입사방향을 모르더라도 정상 동작이 가능한 블라인드 빔 형성 기법 중에서 신호의 cyclostationarity를 이용한 방법을 사용하여 원하는 사용자 신호의 cyclic 주파수만큼 천이된 신호와 수신신호 사이의 상호상관 행렬을 구하고 이를 이용하여 계수벡터를 구하는 알고리즘을 제시한다.

• Proceedings of the Acoustical Society of Korea Conference
• /
• autumn
• /
• pp.201-204
• /
• 2000

음성신호에 섞인 잡음을 처리하기 위해서 단 일 마이크로폰을 이용한 방법이 많이 연구되고 있는데, 그 중에서 Kalman 필터를 이용한 방법은 먼저 음성신호의 모델을 검출하고 잡음이 섞인 신호에서 표준 Kalman 필터를 이용해서 음성신호 성분만을 검출하게 된다. 본 논문에서는 음성신호에 섞인 유색잡음을 백색화하는 방법을 적용하여 Kalman 필터의 잡음제거 성능을 향상시키는 방법을 제안하였다.

• Journal of Korea Multimedia Society
• /
• v.10 no.8
• /
• pp.1004-1016
• /
• 2007

In this paper we proposed a Weight-Update-Control Adaptive Noise Canceller which improves speech when environmental noise is stationary and it is hard to acquire a reference signal. Adaptive Noise Canceller(ANC) needs a reference signal, but it is not easy to measure pure noise without voice for reference in factory. Because there are mixed various mechanical noise and workers' voice. Therefore ANC is not suitable to reduce background noise. So we proposed the method that uses an arbitrary constant as an input signal and inputs microphone signal to the reference signal. The noise is eliminated using updated weights in non-speech range. In speech range the weight is fixed and the modified voice is acquired then voice is restored through transversal filter. The proposed method is based on facts that the factory noise is stationary and the noise is not changed in short conversation range. As a result of simulation using MATLAB, we confirmed that the proposed method is effective for reducing factory noise and has high signal to noise ratio(SNR).

• Annual Conference on Human and Language Technology
• /
• 2009.10a
• /
• pp.81-86
• /
• 2009

비수지 신호는 수화를 통한 의사소통을 하는 과정에 수지 신호 못지않게 중요한 정보를 제공한다. 그럼에도 불구하고 비수지 신호에 대한 연구는 수지 신호에 대한 연구에 비해 아직까지 매우 부족하다. 본 연구에서는 이러한 비수지 신호의 특징에 대해 분석하였다. 비수지 신호를 수지 신호와 함께 재현하는 과정에는 정확도 문제와 동기화 문제가 있는데, 본 연구에서는 동기화 문제를 해결하는 시스템을 제안한다. 구현된 시스템은 입력된 문장을 구문 분석하여 수지 신호와 비수지 신호를 결정하는 부분과 구문 분석된 결과를 바탕으로 수화 애니메이션을 위한 액션 스크립트를 생성하는 부분으로 나뉜다. 수지 신호와 비수지 신호의 연결 순서와 방식에 따라 수화의 뜻이 달라질 수 있기 때문에, 본 연구에서 다루는 비수지 신호의 동기화 문제는 수화 자동 생성에 있어서 매우 중요한 문제이다.

• Investigative Magnetic Resonance Imaging
• /
• v.6 no.2
• /
• pp.158-165
• /
• 2002

Purpose : To evaluate the effect of Gd-DTPA on signal intensity of diffusion-weighted magnetic resonance(MR) image and apparent diffuse coefficient (ADC) in dog brain with hype racute stroke. Materials and methods : Experimental canine model of hyperacute cerebral infarction was made by selective intraarterial embolization with particulate embolic material. Diffusion-weighted MR imaging was performed in five dogs at 1 hour after the embolization of internal carotid artery. After intravenous bolus injection of Gd- DTPA, additional 11 diffusion-weighted MR images were serially obtained from 2 minutes to 90 minutes after injection in each dog. The author evaluated findings of hyperacute cerebral infarction on diffusion-weighted MR imaging, and calculated mean signal intensity and mean ADC in infarcted region and contralateral normal region. Statistical analysis of mean signal intensity, mean ADC and contrast-noise ratio before and after Gd-DTPA injection was performed. Results : Hyperacute cerebral infarction developed in all five dogs on diffusion-weighted MR images obtained 1 hour after embolization. The area of hyperacute infarction had steady increase in signal intensity on diffusion-weighted MR image and decrease in ADC. In normal perfusion area, decrease in signal intensity was observed at 2 minutes the Gd-DTPA injection, whereas ADC did not changed. Conclusion : Intravenous injection of Gd-DTPA had no influence on ADC in both hyperacute infarction and normally perfused are a, but caused initial transient signal reduction in normally perfused area on diffusion-weighted MR image due to susceptibility effect of Gd-DTPA. It is important to calculate ADC in evaluating the effect of diffusion after injection of Gd-DTPA.

• The Journal of the Acoustical Society of Korea
• /
• v.28 no.6
• /
• pp.520-525
• /
• 2009

The conventional studies about an adaptive beamformer assumed that the interference signals are stationary, so they used time-average of signals or Least Mean Squares. However, these methods showed low performance of canceling the non-stationary interferences. In this paper, the MAFF-RLS algorithm is developed in order to cancel non-stationary interferences, and the GSC structure using this algorithm is proposed. Furthermore, the performance of the MAFF-RLS beamformer is verified by simulation using MATLAB. This simulation results show the performance of the proposed beamformer is better than that of the SMI and the conventional RLS beamformer.