• The Journal of the Acoustical Society of Korea
• /
• 제20권4E호
• /
• pp.31-38
• /
• 2001

A large-scale synthesis database for a unit selection based synthesis method usually retains redundant synthesis unit instances, which are useless to the synthetic speech quality. In this paper, to eliminate those instances from the synthesis database, we proposed a new pruning method called weighted vector quantization (WVQ). The WVQ reflects relative importance of each synthesis unit instance when clustering the similar instances using vector quantization (VQ) technique. The proposed method was compared with two conventional pruning methods through the objective and subjective evaluations of the synthetic speech quality: one to simply limit maximum number of instance, and the other based on normal VQ-based clustering. The proposed method showed the best performance under 50% reduction rates. Over 50% of reduction rates, the synthetic speech quality is not seriously but perceptibly degraded. Using the proposed method, the synthesis database can be efficiently reduced without serious degradation of the synthetic speech quality.

• 한국정보디스플레이학회:학술대회논문집
• /
• 한국정보디스플레이학회 2006년도 6th International Meeting on Information Display
• /
• pp.215-217
• /
• 2006

This paper presents an area efficient 8-bit current digital to analog convector (DAC) which is applied to 240 channels Active Matrix - Organic Light Emitting Diode (AMOLED) data driver. The proposed circuit constitutes 4-bit binary weighted current DAC and 4-bit switched capacitor cyclic DAC. The proposed DAC has about 70% smaller area than that of the typical binary weighted current DAC. We overcome sampling time by reducing the number of repetition phases so that it can display 8-bit gray scale image.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2003년도 하계학술대회 논문집 D
• /
• pp.2088-2090
• /
• 2003

A practical recursive weighted least square(WLS) solution is proposed to solve the passive ranging problem. Apart from the previous works based on the extended Kalman filter(EKF), to ensure the convergency at long-range, the proposed scheme makes use of line-of-sight(LOS) rate instead of bearing information. The influence of LOS rate measurement errors is investigated and it is asserted that the WLS estimates contain bias and scale factor errors. Together with simple compensation algorithm, the estimation errors of proposed filter can be reduced dramatically.

• 한국콘텐츠학회:학술대회논문집
• /
• 한국콘텐츠학회 2019년도 춘계종합학술대회
• /
• pp.211-212
• /
• 2019

The objectives of this research are to test the utility of semiparametric geographically weighted regression (SGWR, a spatial analysis method) in the small-scale urban sample, and to understand the geographic patterns of provision and pricing of sharing economy based accommodations in the tourist city. This paper focused on how network distance to heritage site, to casino, residential unit prices and other five attribute categories determine Airbnb price in Macau SAR, China. Findings show that SGWR models outperformed OLS models. Moreover, comparing with heritage sites, casinos are the stronger factors to drive up Airbnb (including hostels) rooms' provision and their prices; and residential unit prices are not related with the Airbnb price in the attraction clusters in Macau. This research showed a little example for the applications of SGWR in the small city, and for the analysis of online marketplace data as new urban study material. Practically, this study provides some scientific evidence for hosts, guests, urban planners, and policymakers' decision making in Macau.

• Advances in nano research
• /
• 제7권2호
• /
• pp.99-108
• /
• 2019

Higher-order theories are very important to investigate the mechanical properties and behaviors of nanoscale structures. In this study, a free vibration behavior of SiNW resting on elastic foundation is investigated via Eringen's nonlocal elasticity theory. Silicon Nanowire (SiNW) is modeled as simply supported both ends and clamped-free Euler-Bernoulli beam. Pasternak two-parameter elastic foundation model is used as foundation. Finite element formulation is obtained nonlocal Euler-Bernoulli beam theory. First, shape function of the Euler-Bernoulli beam is gained and then Galerkin weighted residual method is applied to the governing equations to obtain the stiffness and mass matrices including the foundation parameters and small scale parameter. Frequency values of SiNW is examined according to foundation and small scale parameters and the results are given by tables and graphs. The effects of small scale parameter, boundary conditions, foundation parameters on frequencies are investigated.

• Pediatric Gastroenterology, Hepatology & Nutrition
• /
• 제24권1호
• /
• pp.38-44
• /
• 2021

Purpose: Digital communication is becoming increasingly important in clinical practice and research. The finding that stool consistency can be evaluated similarly using either "in vivo" or photographic material by health care professionals will decrease subjective interpretation by parents. The primary outcome of this study was the reliability of stool consistency scoring using the Brussels Infant and Toddler Stool Scale (BITSS) between fresh stools and their photos; the secondary outcome was the inter-rater reliability based on the fresh stools. Methods: Fresh stool samples from healthy children were collected in a day care center. These stools, and one month later the corresponding photos presented in a random order, were presented to 14 observers. Reliabilities were analyzed using absolute agreements and weighted and unweighted Cohen's κ. Results: In total, 202 samples were rated 576 times. Absolute agreement between photographic and real time assessment ranged between 71.1% and 83.3% among observers. This corresponded with substantial agreement (unweighted κ=0.70 [95% CI, 0.61-0.78]; weighted κ=0.86 [95% CI, 0.78-0.88]). The inter-observer agreement showed similar percentages of absolute agreement (81.4-82.0%) and κ-values corresponding with fair-to-moderate agreement. Conclusion: Our findings suggest that the assessment of fresh stool consistency can also reliably be done on photographic material when using the BITSS. This opens opportunities in scientific surroundings and in our daily life communication with parents and caretakers.

• 대한의용생체공학회:의공학회지
• /
• 제21권4호
• /
• pp.333-338
• /
• 2000

본 논문에서는 T1 강조영상, T2 강조 영상 그리고 PD의 영상의 특징을 상호 보완적으로 이용한 자동적인 영상 분할법을 제안한다. 제안한 분할 알고리듬은 3단계로 이루어지는데, 첫 단계에서는 PD 영상으로부터 대뇌 마스크를 획득한 후, T1과 T2, PD의 입력 영상에 대뇌 마스크를 씌워 각각의 대뇌 영상을 추출하고, 둘째 단계에서는 대뇌 내부 조직에 해당하는 두드러진 클러스터(outstanding cluster)를 3차원 클러스터들 중에서 선택한다. 3차원 클러스터는 최적스케일 영상(optimal scale image)으로 이루어지는 3차원 공간상에서 화소가 밀집된 봉우리들을 교집합해서 생성되는 클러스터로 결정한다. 최적스케일 영상은 각 2타원 히스토그램에 스케일 스페이스 필터링을 적용시키고 그래프(graph) 구조를 검색하여 2차원 히스토그램의 모양을 가장 잘 나타내는 봉우리(peak) 영상을 최적 스케일 영상으로 선택한다. 마지막 단계에서는 앞에서 찾은 두드러진 클러스터의 중심값을 FCM 알고리듬의 초기중심 값으로 두고, FCM 알고리듬을 이용하여 대뇌 영상을 분할한다. 제안한 분할 알고리듬은 정확한 클러스터의 중심값을 계산함으로 초기 값을 영향을 많이 받는 FCM 알고리듬의 단점을 보완하였고 다중 스펙트럼 영상의 특성을 조합하여 분할에 이용함으로 단일 스펙트럼 영상만을 이용하는 방법보다 향상된 결과를 얻을 수 있었다.

• 한국지리정보학회지
• /
• 제23권3호
• /
• pp.263-274
• /
• 2020

한국 남동해역은 매년 하계에 집중적으로 냉수대가 형성되어 빈번한 이상해황이 발생한다. 본 연구에서는 이 해역에서 발생하는 표층수온 분포의 공간 변화를 분석하기 위해 2018년 6월에서 9월까지 고리와 정자 부이에서 관측한 해양현장 수온 데이터와 GHRSST Level 4 재분석 해수면 온도(sea surface temperature: SST) 자료를 이용하였다. 부이 자료는 두 지점의 시계열적 수온 변동 분석에, GHRSST 자료는 연구해역 전반에 걸친 일별 SST의 분산과 가중공간중심(weighted mean center: WMC)을 계산하는데 이용하였다. 부이의 수온이 낮아지면 연구해역 SST의 분산이 증가하는 경향을 보였으나, 전 기간 일치하게 나타나지는 않았다. 이는 GHRSST가 재분석 자료로 연안의 민감한 수온변화를 반영하지 못하기 때문이다. 이와 같이 전 해역의 SST 변화를 대표하는 통계적 분산만으로는 연안의 국지적인 소규모의 수온변화를 파악하거나, 냉수대 발생해역의 위치 및 범위를 탐지하기에는 한계가 있다. 따라서 차가운 수괴가 발생하는 공간적인 위치를 정량적으로 파악하기 위해 WMC를 활용하여 분석한 결과 냉수대가 발생했을 때, WMC가 연구해역의 공간중심(mean center: MC)으로부터 북서 해역 쪽에 위치하였다. 이는 SST의 WMC 위치 정보를 통해 차가운 표층수온의 분포가 어디에서 어느 정도 나타나는지를 정량적으로 파악할 수 있음을 의미하며, 향후 냉수대 규모 및 지역 확산 범위 탐지에 WMC의 활용 가능성을 알 수 있었다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제11권10호
• /
• pp.4929-4947
• /
• 2017

In visual object tracking, Correlation Filter-based Tracking (CFT) systems have arouse recently to be the most accurate and efficient methods. The CFT's circularly shifts the larger search window to find most likely position of the target. The need of larger search window to cover both background and object make an algorithm sensitive to the background and the target occlusions. Further, the use of fixed-sized windows for training makes them incapable to handle scale variations during tracking. To address these problems, we propose two layer target representation in which both global and local appearances of the target is considered. Multiple local patches in the local layer provide robustness to the background changes and the target occlusion. The target representation is enhanced by employing additional reversed RGB channels to prevent the loss of black objects in background during tracking. The final target position is obtained by the adaptive weighted average of confidence maps from global and local layers. Furthermore, the target scale variation in tracking is handled by the statistical model, which is governed by adaptive constraints to ensure reliability and accuracy in scale estimation. The proposed structural enhancement is tested on VTBv1.0 benchmark for its accuracy and robustness.

• 대한기계학회논문집
• /
• 제11권3호
• /
• pp.528-536
• /
• 1987

본 연구에서는 간헐적인 유동에 의거한 체적 대류 모델을 설정하고, 구배 확 산 모델에 유동장의 중심부와 외부에서 서로 다른 무게값을 부여하는 혼성 확산 모델 (hybrid diffusion model)을 제안하며, 이 모델을 검증하기 위한 첫 단계로서 평면 제 트 유동에 대하여 수치 계산을 수행하다. 여기에는 간헐도에 대한 난류 전달 방정식 이 필요한데 이 방정식의 생성항(production term)은 곧 외부의 비회전 유동이 난류성 유체로 유입되는 정도를 의미하게 된다.