• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.7 no.5
• /
• pp.907-912
• /
• 2006

Pre-emphasizing the speech compensates for falloff at high frequencies. The most common form of pre-emphasis is y(n)=s(n)-A${\cdot}$s(n-1), where A typically lies between 0.9 and 1.0 in voiced signal. And, this value reflects the degree of pre-emphasis and equals R(1)/R(0) in conventional method. This paper proposes a new flattening method to compensate the weaked high frequency components that occur by vocal cord characteristic. We used QMF(Quardrature Mirror Filter) to minimize the output signal distortion. After using the QMF to compensate high frequency components, flattening process is followed by R(1)/R(0) at each frame. Experimental results show that the proposed method flattened the weaked high frequency components effectively than auto correlation method. Therefore, the flattening algorithm will apply in speech signal processing like speech recognition, speech analysis and synthesis.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.11 no.9
• /
• pp.3558-3562
• /
• 2010

The exact fundamental frequency (pitch) extraction is important in speech signal processing. However the exact pitch extraction from speech signal is very difficult due to the effect of formant and transitional amplitude. So in this paper, the pitch is detected after flattening the spectrum in frequency region by proposed algorithm for minimized harmonics variance value. Experimental result showed the proposed method appeared an outstanding performance in compared with LPC, Cepstrum. Also, the results show the proposed method is better than conventional method.

• KIISE Transactions on Computing Practices
• /
• v.21 no.12
• /
• pp.780-785
• /
• 2015

Recently HTTP (Hypertext Transfer Protocol) adaptive streaming services have been the subject of much attention. The video quality control scheme of conventional HTTP adaptive streaming services estimates bandwidth using segment throughput and smooths out the sample of segment throughput. However, the conventional scheme has the problem of QoE (Quality of experience) degradation occurring with buffer underflow and frequent quality change due to the fixed number of samples. In order to solve this problem, we propose a video quality control scheme based on segment throughput and buffer occupancy. The proposed scheme determines the number of samples according to the variation of segment throughput. The proposed scheme also controls video quality based on the threshold of bitrate to keep stable buffer occupancy. The simulation results show that proposed scheme improves QoE by preventing buffer underflow and decreasing quality change when compared with the conventional scheme.

• The Journal of the Korea Contents Association
• /
• v.3 no.3
• /
• pp.55-62
• /
• 2003

In this paper, I proposed a new WRR Cell Scheduling algorithm for improving the problem of existing smoothing scheme. in proposed algorithm, when queue of being service VC in present round is emptied, instead VC that difference between assigned weight and length of present buffer is biggest in all VC service. This result efficiency of bandwidth allocation is increased. And above all real-time traffic is serviced, non-real-time to allocate bandwidth of the remainder multimedia-service is possible. In one cycle, number of serviced average cell is increased. And for many eel serviced, length of total Duffer is decreased.

• Proceedings of the IEEK Conference
• /
• 2003.07e
• /
• pp.1715-1718
• /
• 2003

본 논문에서 우리는 정규화 된 혼합 노름(norm)을 이용한 다중 채널 영상 복원 알고리즘을 제안한다. 채널 내부와 채널 사이의 결정론적 정보를 이용하는 다중채널 복원 문제를 고려한다. 각 채널에서, LMS(Least Mean Square), LMF(Least Mean Fourth), 평탄 함수가 결합된 함수가 제안되었다. LMS와 LMF 사이의 적절한 분배를 제어하는 혼합 노를 매개변수와 해의 평탄 정도를 정의하는 정규화 매개 변수를 소개하며, 두 매개 변수는 각 채널의 잡음 특성에 따라 매번 반복적으로 갱신된다. 제안된 알고리즘은 각 채널의 잡음분포에 대한 지식이 필요하지 앉고 앞에서 언급된 매개 변수는 부분적으로 복원된 영상에 기반을 두고 조절하게 된다.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.23 no.6
• /
• pp.423-438
• /
• 2021

The blade of motor grader is used for scattering and leveling the aggregates on the foundation of road construction site. The paper performed a design improvement research of the blade part to enhance the working efficiency of motor graders. The scattering works of aggregates by blade driving were simulated by DEM (discrete element method) of a dynamic code. The four design parameters were selected and a specific leveling scenario for the simulation was determined. The nine blade models were numerically experimented, and the sensitivity of each factors was analyzed. Next, the design factors that influence a blade performance have been selected by ANOVA, and these key design factors were applied to the progressive quadratic response surface method (PQRSM). The optimum set of design factors of the blade was finally proposed.

• The Journal of Bigdata
• /
• v.4 no.2
• /
• pp.61-71
• /
• 2019

Chemical Mechanical Planarization (CMP) process that planarizes semiconductor wafer's surface by polishing is difficult to manage reliably since it is under various chemicals and physical machinery. In CMP process, Material Removal Rate (MRR) is often used for a quality indicator, and it is important to predict MRR in managing CMP process stably. In this study, we introduce prediction models using machine learning techniques of analyzing time-series sensor data collected in CMP process, and the classification models that are used to interpret process quality conditions. In addition, we find meaningful variables affecting process quality and explain process variables' conditions to keep process quality high by analyzing classification result.

• The Journal of the Acoustical Society of Korea
• /
• v.27 no.8
• /
• pp.418-426
• /
• 2008

In this paper, we propose an algorithm for the frequency channel segmentation using peaks and valleys in spectrogram. The frequency channel segments means that local groups of channels in frequency domain that could be arisen from the same sound source. The proposed algorithm is based on the smoothed spectrum of the input sound. Peaks and valleys in the smoothed spectrum are used to determine centers and boundaries of segments, respectively. To evaluate a suitableness of the proposed segmentation algorithm before that the grouping stage is applied, we compare the synthesized results using ideal mask with that of proposed algorithm. Simulations are performed with mixed speech signals with narrow band noises, wide band noises and other speech signals.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2000.04a
• /
• pp.884-889
• /
• 2000

본 논문에서는 원영상의 주파수특성 성형에 의한 오차확산 영상이진화 기법을 제안하였다. 오차확산에 의한 영상이진화 처리에서 이진화영상에 내재하고 있는 에지정보의 열화를 감쇄하기 위한 방법으로 원영상의 주파수특성을 활용하는 기법을 도입하였다. 주파수특성은 원영상에서 수평방향과 수직방향으로 얻고, 이들 각각에 가중치를 부여하였다. 수평과 수직의 주파수특성을 성형한후, 영상이진화 기법에서 양자화처리 수행시 이를 활용하였다. 주파수특성 성형을 위한 필터는 2종을 제안하여 적용하였다. 제안한 필터를 적용한 오차확산 영상이진화 처리를 수행한 후, 실험결과를 고찰시 이진화영상의 평탄영역은 기본특성을 유지하면서 에지정보의 특성이 부각되는 주관적인 화질특성이 얻어짐을 확인할 수 있었다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.11a
• /
• pp.597-598
• /
• 2023

X-ray는 촬영 방식의 한계로 진단하기 어려운, 품질 낮은 영상을 다수 발생시킨다. 이러한 저품질 영상은 임상 현장에서의 진단이 어려울 뿐만 아니라, 진단 보조 도구를 개발함에 모델의 성능과 신뢰도를 떨어뜨리는 주요 요소가 된다. 특히 투과도가 낮은 영상은 학습 성능에 악영향을 미친다는 것이 입증된 바 있다. 따라서 본 연구는 투과도가 낮은 영상을 진단에 부적합한 영상으로 정의하여, 이를 분류하는 기법을 제안한다. 제안하는 기법은 민감도 94.9%. 특이도 96.0%의 높은 성능을 보였다.