• Proceedings of the Acoustical Society of Korea Conference
• /
• spring
• /
• pp.107-110
• /
• 2000

음성 신호는 시간에 따라 변하지만 일정 구간에서는 특성이 변하지 않는다고 가정하여 윈도우를 취해 단구간 분석을 한다. 이 때 윈도우의 적용은 필수적이다. 하지만 단구간 분석을 위해서 사용되는 윈도우에 의해 생기는 누설에너지 때문에 음성신호의 스펙트럼 정보가 왜곡되어 버린다. 본 논문에서는 스펙트럼 분석 시 발생되는 누설에너지를 최소화하는 방법을 제안하고자 한다. 음성신호에 고정된 크기의 rectangular Window를 취한 후 처음 샘플과 차이가 가장 작은 샘플을 프레임 크기의 3/4인 지점에서부터 검색하여 최소인 부분까지 시간 축 스케일링을 한 후 기존의 윈도우 크기와 같은 크기로 만든다. interpolation과 decimation을 조합하여 시간 축 스케일링을 한다. 이렇게 윈도우가 적용된 신호를 처리 후 위의 역 과정을 수행한다. 제안한 윈도우의 SNRseg는 rectangular window보다 평균 7.88dB 낮고, kaiser window보다 평균 1.65dB 높았다. 또한 제안한 윈도우의 SD는 rectangular window 보다 평균 $1.73dB^\2$ 낮았다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.05a
• /
• pp.298-300
• /
• 2023

자율주행 및 robot navigation 시스템에서 물체 인식 성능향상을 위해 대부분 MSF(Multi-Sensor Fusion) 기반 설계를 한다. 따라서 각 센서로부터 들어온 정보를 정합하는 것은 정확한 MSF 알고리즘을 위한 필요조건이다. 다양한 선행 연구에서 2D 데이터에 대한 공격을 진행했다. 자율주행에서는 3D 데이터를 다루어야 하므로 선행 연구에서 하지 않았던 3D 데이터 공격을 진행했다. 본 연구에서는 스케일링 공격 기반 카메라-라이다 센서 간 정합 모델의 정확도를 저하시키는 공격 방법을 제안한다. 제안 방법은 입력 라이다의 포인트 클라우드에 스케일링 공격을 적용하여 다운스케일링 단계에서 공격하고자 한다. 실험 결과, 입력 데이터에 공격하였을 때 공격 전보다 평균제곱 이동오류는 56% 이상, 평균 사원수 각도 오류는 98% 이상 증가했음을 보였다. 다운스케일링 크기 별, 알고리즘별 공격을 적용했을 때, 10×20 크기로 다운스케일링 하고 lanczos4 알고리즘을 적용했을 때 가장 효과적으로 공격할 수 있음을 확인했다.

• Journal of the Microelectronics and Packaging Society
• /
• v.20 no.1
• /
• pp.1-5
• /
• 2013

Wafer-to-Wafer (W2W) integration technology is an emerging technology promising many benefits, such as reduced size, improved performance, reduced power, lower cost, and divergent integration. As the maturity of W2W technology progresses, new applications will become more viable. However, at present the cost for W2W integration is still very high and both manufacturing yield and reliability issues have not been resolved yet for high volume manufacturing (HVM). Especially for WTW integration resolving compound yield issue can be a key factor for HVM. To have the full benefits of WTW integration technology more than simple wafer stacking technologies are necessary. In this paper, the manufacturing yield for W2W integration is described and the challenges of WTW integration will be discussed.

• Journal of Environmental Science International
• /
• v.29 no.1
• /
• pp.25-32
• /
• 2020

The determination of soil parameters is important in predicting the simulated surface runoff using either a distributed or a lumped rainfall-runoff model. Soil characteristics can be collected using remote sensing techniques and represented as a digital map. There is no universal agreement with respect to the determination of a representative parameter from a gridded digital map. Two representative methods, i.e., arithmetic and predominant, are introduced and applied to both FLO-2D and HEC-HMS to improve the model's accuracy. Both methods are implemented in the Yongdam catchment, and the results show that the former seems to be more accurate than the latter in the test site. This is attributed to the high conductivity of the dominant soil class, which is A type.

• Journal of Korea Multimedia Society
• /
• v.8 no.10
• /
• pp.1329-1336
• /
• 2005

We propose an optimal edge detection algorithm adaptive to the various widths of edges. In order to construct this algorithm, we introduce an alternative definition of edge point and generalize the directional derivatives in the pixel space to obtain an extended directional derivatives beyond scaling. The result from applying our algorithm to a 2D image is analyzed comparing to that from the other algorithm.

• Journal of Broadcast Engineering
• /
• v.25 no.2
• /
• pp.242-251
• /
• 2020

Recently, many deep convolutional neural networks for image super-resolution have been studied. Existing deep learning-based super-resolution algorithms are architecture that up-samples the resolution at the end of the network. The post-upsampling architecture has an inefficient structure at large scaling factor result of predicting a lot of information for mapping from low-resolution to high-resolution at once. In this paper, we propose a single image super-resolution using Channel Attention Residual Dense Block based on an iterative up-down sampling architecture. The proposed algorithm efficiently predicts the mapping relationship between low-resolution and high-resolution, and shows up to 0.14dB performance improvement and enhanced subjective image quality compared to the existing algorithm at large scaling factor result.

• Journal of Korea Multimedia Society
• /
• v.9 no.9
• /
• pp.1160-1172
• /
• 2006

Unfolding is an efficient technique to detect colorectal disease since it represents entire shape of colon at a glance. However, the previous unfolding methods may miss the surface located behind the folds or high curvature areas. To solve the problem, we propose an adaptive unfolding method that controls the number of rays in accordance with the surface complexity. In the first step, we find sample points of longer and shorter distance between the sample points and the colon surface, and then determine them as based control points. Next, for reaching rays of sample points in the wrinkled colon surface, sample points are shifted toward the control points that have the long distance. Lastly, we perform a 2D scaling which solves local distortions of the unfolding surface and avoids missing surface. Consequently, our method could be visualized the hidden surface of colon organ in detail.

• Journal of The Korean Astronomical Society
• /
• v.49 no.5
• /
• pp.193-198
• /
• 2016

Early-type galaxies (ETGs) are supposed to follow the virial relation $M=k_e{\sigma}^2R_e/G$, with M being the mass, σ_* being the stellar velocity dispersion, R_e being the effective radius, G being Newton's constant, and k_e being the virial factor, a geometry factor of order unity. Applying this relation to (a) the ATLAS^3D sample of Cappellari et al. (2013) and (b) the sample of Saglia et al. (2016) gives ensemble-averaged factors 〈k_e〉 = 5.15 ± 0.09 and 〈k_e〉 = 4.01 ± 0.18, respectively, with the difference arising from different definitions of effective velocity dispersions. The two datasets reveal a statistically significant tilt of the empirical relation relative to the theoretical virial relation such that $M{\propto}({\sigma}^2_*R_e)^{0.92}$. This tilt disappears when replacing R_e with the semi-major axis of the projected half-light ellipse, a. All best-fit scaling relations show zero intrinsic scatter, implying that the mass plane of ETGs is fully determined by the virial relation. Whenever a comparison is possible, my results are consistent with, and confirm, the results by Cappellari et al. (2013). The difference between the relations using either a or R_e arises from a known lack of highly elliptical high-mass galaxies; this leads to a scaling (1 - ϵ ) ∝ M^0.12, with ϵ being the ellipticity and $R_e=a\sqrt[]{1-{\epsilon}}$. Accordingly, a, not R_e, is the correct proxy for the scale radius of ETGs. By geometry, this implies that early-type galaxies are axisymmetric and oblate in general, in agreement with published results from modeling based on kinematics and light distributions.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.12 no.7
• /
• pp.1191-1198
• /
• 2001

MMIC low noise amplifiers for millimeter-wave application using 0.15 $\mu$m pHEMT have been presented in this paper. The design emphasis is on active device model and EM simulation. The deficiency of conventional device models is identified. A distributed device model has been adapted to circumvent the scaling problems and, thus, to predict small signal and noise parameters accurately. Two single-ended low noise amplifier are designed using distributed active device model for Q-band(40 ∼ 44 GHz) and V-band(58 ∼65 GHz) application. The Q-band amplifier achieved a average noise figure of 2.2 dB with 18.3 dB average gain. The V-band amplifier achieved a average noise figure of 2.9 dB with 14.7 dB average gain. The design technique and model employed provides good agreement between measured and predicted results. Compared with the published data, this work also represents state-of-the-art performance in terms of gain and noise figure.

• Journal of Ocean Engineering and Technology
• /
• v.35 no.1
• /
• pp.75-81
• /
• 2021

Artificial Coral Reefs (ACRs) have been introduced to help solve coastal erosion problems, but their feasibility has not been assessed with field data. This study conducted a feasibility analysis of ACRs on their erosion mitigation effects by performing a case study of Cheonjin-Bongpo beach, South Korea. A numerical-physical combined analysis was carried out using a SWAN model simulation and physical model test with a scale of 1/25 based on field observations of Cheonjin-Bongpo beach. Both Dean's parameter and the surf-scaling parameter were applied to comparative analysis between the absence and presence conditions of the ACR. The results for this combined method indicate that ACR attenuates the wave height significantly (59~71%). Furthermore, ACR helps decrease the mass flux (~50%), undertow (~80%), and maximum wave set up (~61%). The decreases in Dean's parameter (~66%) and the surf-scaling parameter suggest that the wave properties changed from the dissipative type to the reflective type even under high wave conditions. Consequently, an ACR can enhance shoreline stability.