• The Journal of the Acoustical Society of Korea
• /
• v.21 no.6
• /
• pp.552-558
• /
• 2002

Recently, based on the human speech recognition (HSR) model of Fletcher, the multi-band speech recognition has been intensively studied by many researchers. As a new automatic speech recognition (ASR) technique, the multi-band speech recognition splits the frequency domain into several sub-bands and recognizes each sub-band independently. The likelihood scores of sub-bands are weighted according to reliabilities of sub-bands and re-combined to make a final decision. This approach is known to be robust under noisy environments. When the noise is stationary a sub-band SNR can be estimated using the noise information in non-speech interval. However, if the noise is non-stationary it is not feasible to obtain the sub-band SNR. This paper proposes the inverse sub-band distance (ISD) weighting, where a distance of each sub-band is calculated by a stochastic matching of input feature vectors and hidden Markov models. The inverse distance is used as a sub-band weight. Experiments on 1500∼1800㎐ band-limited white noise and classical guitar sound revealed that the proposed method could represent the sub-band reliability effectively and improve the performance under both stationary and non-stationary band-limited noise environments.

• Proceedings of the IEEK Conference
• /
• 1998.06a
• /
• pp.631-634
• /
• 1998

본 논문은 음향반향제거기에서 동시통화구간을 검출하고 처리하는 알고리듬에 대하여 논하고 있다. 새로운 동시통화처리 알고리듬은 송신신호와 잔류신호간의 상호 상관계수를 이용하여 동시통화 여부 및 구간을 검출하고 이 구간에서는 반향제거기의 계수적응을 중지하도록 하는 것이다. 본 알고리듬을 사용함으로서 동시통화구간의 검출에 소요되는 계산량이 대폭 감소할 뿐만 아니라 비정상상태의 원인이 동시통화에 의한 것인지 아니면 반항경로응답의 변화에 의한 것인지를 구별할 수가 있다. 모의실험을 수행하여 본 알고리듬과 기존의 것과의 성능을 비교, 검토하였다. 실험 결과 몬 알고리듬을 적용하므로서 동시통화 후에도 계수가 발산하지한ㅎ고 안정적으로 정상상태를 유지함을 확인하였다.

• The Journal of the Acoustical Society of Korea
• /
• v.17 no.3
• /
• pp.10-16
• /
• 1998

본 논문은 음향반향제거기에서 동시통화구간을 검출하고 처리하는 알고리듬에 대하 여 논하고 있다. 새로운 동시통화처리 알고리듬은 송신신호와 잔류신호간의 상호상관계수를 이용하여 동시통화 여부 및 구간을 검출하고 이 구간에서는 반향제거기의 계수적응을 중지 하도록 하는 것이다. 본 알고리듬을 사용함으로서 동시통화구간의 검출에 소요되는 계산량 이 대폭 감소할 뿐만 아니라 비정상상태의 원인이 동시통화에 의한 것인지 아니면 반향경로 응답의 변화에 의한 것인지를 구별할 수가 있다. 음향반향제거기에 제안한 알고리듬을 적용 하여 모의실험을 수행하므로서 본 알고리듬의 성능을 분석하였고 기존의 알고리듬을 적용한 경우와 비교, 검토하였다. 실험 결과 본 알고리듬을 적용한 음향반향제거기는 동시통화구간 을 완벽하게 검출하므로서 동시통화 후에도 계수가 발산하지않고 안정적으로 정상상태를 유 지함을 확인하였다.

• Proceedings of the Korea Institute of Convergence Signal Processing
• /
• 2003.06a
• /
• pp.1-5
• /
• 2003

의료영상의 분할은 의료영상을 컴퓨터 진단 및 가시화에 필요한 같은 성질을 가진 여러 조직으로 나누어주는 방법이다. 즉 입력되어진 영상을 처리하여 유사한 화소들의 집합인 영역들로 화소들을 구분하는 작업이며 영상분할의 결과는 영상인식의 정확성에 큰 영향을 미친다. MRI(Magnetic Resonance Imaging)으로부터 정상적인 세포조직 또는 뇌종양과 같은 비정상적인 세포조직의 가시화와 분석을 위해서는 대상 세포조직의 적절한 분류를 필요로 한다. 하지만 기존의 영역 검출 방법으로는 잡음이 섞여 있는 영상에서 여러 가지의 처리과정(주로 잡음 제거)이 필수적이고 그런 과정으로 인해 정확한 영역 검출이 힘들게 된다. 이에 잡음이 있더라도 이를 제거하기 위한 처리가 필요 없이 영역기반으로 필요한 파라미터의 추정을 통한 MRF(Markov Random Field)를 이용하여 보다 효율적이고 정확하게 MRI에서 질환 영역을 검출할 수 있다.

• Proceedings of the Korea Contents Association Conference
• /
• 2005.11a
• /
• pp.649-653
• /
• 2005

To estimate for an error signal with sigmoid nonlinearity what reduced constellation applies closed eye pattern in the initial equalization, there can be improves problems of previous soft decision-directed algorithm that increasing estimate complexity and decreasing of convergence speed when substitute high-order constellation. The characteristic of sigmoid function is adjusted by a mean and a variance parameter, so it depends on adjustment of variance that what reduced constellation $values(\gamma)$ can have ranges between + $\gamma$ and - $\gamma$. In this paper, we proposed Variable Modulus Algorithm (VMA) that can be improving a performance of steady-state by adjustment of variance when equalization works normally and each cluster of constellation decrease.

• Progress in Medical Physics
• /
• v.7 no.1
• /
• pp.19-23
• /
• 1996

The chemical bond differences between a normal tissue and a fat tissue make a chemical shift artifact which is caused by a primary inacuracy of resonance signal location. The chemical shift also makes a variation of the transverse time T$_2$. An attempt is made to compare the values of SNR(Signal-to-Noise Ratio), the signal response, and the imaging time computed by applying T$\sub$2/$\^$＊/ for a fat-proton with ones of those computed by applying T$_2$ for a water-proton under the conditions of T$_1$/T$_2$=3 and T$\sub$2/$\^$＊/T$_2$=0.9. The results of the attempt show that the first two reduce to 5％ and 8％ out of 100％, respectively, and the last rather increases up to 10％. This shows that the chemical shift contributes to the deterioration of an MR imaging efficiency in addtion to the image distortion.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.13 no.4
• /
• pp.127-133
• /
• 2013

This paper compare the performance of CCA (Compact Constellation Algorithm) adaptive equalization algorithm by effect of the compact slice weighting value for minimization of the intersymbol interference in the communication channel. The CCA combines the conventional DDA and RCA algorithm, it uses the constant modulus of the transmission signal and the considering the output of decision device by the power of compact slice weighting value in order to improving the initial convergence characteristics and the equalization noise by misadjustment in the steady state. In this process, it is confirmed by computer simulation that the compact slice weight affects the performance of CCA adaptive equalization algorithm. The performance index includes the output signal constellation, the residual isi and maximum distortion and MSE that is for the convergence characteristics, the SER according to the signal and noise power ratio at the channel is used. As a result of computer, it shows that the large weighting value gives more good in every performance index. But in SER performance, it is known that the small values gives more good in low SNR and the large values gives more good in high SNR.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.6 no.2
• /
• pp.3-12
• /
• 1996

Sliding mode controller(SMC) is a simple but powerful nonlinear controller, because it guarantees the stability and the robustness. However, it leads to the high frequency chattering of the control input. Although the phenomenon can be avoided by introducing a thin boundary layer to the sliding surface, the method results in a steady state: error proportional to the boundary layer thickness. In this paper, we proposed a new sliding mode controller with self-tuning the thickness of a boundary layer. It uses a fuzzy rule base for tuning the thickness of a boundary layer. That is, the thickness is increased to some degree to reject a discontinuous control input at the initial state and then it is decreased as the states approaches to the steady states for improving the tracking performance. In order to assure the control performance, we perf'ormed the computer simulation using an inverted pendulum system as a controlled plant.

• Journal of the Korean earth science society
• /
• v.30 no.6
• /
• pp.692-698
• /
• 2009

Wavelet transform has been widely used in terms that it may overcome the shortcoming of conventional Fourier transform. Fourier transform has its difficulty to explain how the transformed domain, frequency, is related with time. Traditional semblance technique in Fourier transform was devised to compare two time series on the basis of their phase as a function of frequency. But this method is known not to work well for the non-stationary signal. In this study, we present two applications of the wavelet-based semblance method to geophysical data. Firstly, we show filtered geomagnetic signal remained with components of high correlation to each observatory. Secondly, highly correlated residual signal of gravity and magnetic survey data, which are also filtered by this semblance method, is present.

• Journal of the Institute of Convergence Signal Processing
• /
• v.16 no.3
• /
• pp.90-95
• /
• 2015

This study presents a method to test quite small-sized and light-weighted micro-drill bits which are used to make holes in printed circuit boards(PCB). After getting images of micro-drill bits through the high resolution microscope, we developed image processing algorithms to detect fiducial points, and then measured diverse factors of the drill-bit based on these points. We also developed the window-based inspection system to automatically discriminate normal and abnormal status. For the relative comparison of its performance, the system was compared with an existing inspection system using test images. Experimental results showed that the proposed system slightly improved performance, and also classified correctly some misjudged errors which were occurred in the existing system.