• 제목/요약/키워드: Verification experiment

검색결과 787건 처리시간 0.031초

Voice Similarities between Brothers

  • Ko, Do-Heung;Kang, Sun-Mee
    • 음성과학
    • /
    • 제9권2호
    • /
    • pp.1-11
    • /
    • 2002
  • This paper aims to provide a guideline for modelling speaker identification and speaker verification by comparing voice similarities between brothers. Five pairs of brothers who are believed to have similar voices participated in this experiment. Before conducted in the experiment, perceptual tests were measured if the voices were similar between brothers. The words were measured in both isolation and context, and the subjects were asked to read five times with about three seconds of interval between readings. Recordings were made at natural speed in a quiet room. The data were analyzed in pitch and formant frequencies using CSL (Computerized Speech Lab), PCQuirer and MDVP (Multi -dimensional Voice Program). It was found that data of the initial vowels are much more similar and homogeneous than those of vowels in other position. The acoustic data showed that voice similarities are strikingly high in both pitch and formant frequencies. It was also found that the correlation coefficient was not significant between parameters above.

  • PDF

기계 학습을 이용한 인공지지체 외형 불량 예측 모델에 관한 연구 (A Study on Prediction Model of Scaffold Appearance Defect Using Machine Learning)

  • 이송연;허용정
    • 반도체디스플레이기술학회지
    • /
    • 제19권2호
    • /
    • pp.26-30
    • /
    • 2020
  • In this paper, we studied the problem if the experiment number occurring in order to identify defect in scaffold. We need to change each of the 5 print factor to predict defect when printing disk type scaffold using FDM 3d printer. So then the number of scaffold print will be more than 100,000 times. This experiment number is difficult to perform in the field. In order to solve this problem, we have produced a prediction model based on machine learning multiple linear regression using print conditions and defect scaffold data for print conditions. The prediction model produced was verified through experiments. The verification confirmed that the error was less than 0.5 %. We have confirmed that satisfied within the target margin of error 5 %.

일반적인 형상의 스탬핑의 최적블랭크 설계를 통한 민감도법의 검증 (Verification of Sensitivity Method for the Design of Optimal Blanks of General Shaped Parts)

  • 손기찬;심현보;황현태
    • 소성∙가공
    • /
    • 제10권1호
    • /
    • pp.15-22
    • /
    • 2001
  • The sensitivity method has been utilized to find initial blank shapes which transform into desired shapes after forming. From the information of die shapes, target shape and material properties, the corresponding initial blank which gives final shape after deformation has been found. Drawings of a trapezoidal cup, a cross-shaped cup and an oil pan have been chosen as the examples. At every case the optimal blank shape has been obtained only a few times of modification without any predetermined deformation path. With the predicted optimal blank, both computer simulation and experiment are performed. Excellent agreements are recognized between simulation and experiment at every cases Through the investigation, the sensitivity method is found to be effective in obtaining optimal blank shapes in drawing of complex shapes.

  • PDF

CT Simulation후 DRR film를 이용한 치료중심점을 재확인하기 위한 Conventional simulation의 유용성에 대한 분석

  • 박광호;임영석;김정만
    • 대한방사선치료학회지
    • /
    • 제11권1호
    • /
    • pp.100-105
    • /
    • 1999
  • Purpose : When the value of X,Y,Z coordination of the isocenter are reallocated from an arbitrary point using DRR (Digitally Reconstructed Radiographs) image in CT Simulation, conventional simulation is normally performed to verify the accuracy of this reallocation of the isocenter through the fluroscopy. The purpose of our experiment is to determine whether repeated test of the verification is necessary or not, and to analyze errors of reallocation with respect to the body region and the beam projection, if necessary, Material and Method : For 200 simulation patient, an arbitrary point is marked on each body and axial scaning is performed using CT, and treatment planing is done by drawing tumor and target volume on each slice. Using the planing data and the reallocated point of the isocenter, DRR image can be obtained and the final isocenter are marked on the patient's skin. In order to verify this reallocation of X,Y.Z coordination from CT simulation, We measure and evaluate the errors of these value on the fluoroscopy monitor and systematize them by classifying according to each body region (Brain, Neck and SCL, Lung, Esophagus, abdomen, Breast and Pelvis) and each beam projection {AP(PA), Supine, Prone and conformal : etc. } Conclusion : Isocenters are shifted by 3-5 mm in the case of Neck & SCL, Breast. at Abdomen, while noticeable differences are not found in other regions. Also, there are not correlations between the errors and the body regions or beam projections. However, our experiment intends to decide whether the procedure of verification is necessary on the vase of time and economy. It is regretful that we could not fully analyze the geometrical errors of DRR image and visual errors from the divergence. In conclusion, according to how much doctor consider tumor margin in drawing tumor and target volume, the meaning of analysis on the reallocation of isocenter should be reinterpreted, (which depends on the experience and capability of doctors)

  • PDF

광류를 이용한 영상기반 무인항공기의 자세 추정 (Attitudes Estimation for the Vision-based UAV using Optical Flow)

  • 조선영;김종훈;김정호;조겸래;이대우
    • 한국항공우주학회지
    • /
    • 제38권4호
    • /
    • pp.342-351
    • /
    • 2010
  • UAV는 임무 수행을 위한 INS 장비와 광학 장비를 갖추고 있다. 이 논문에서는 UAV를 위한 알고리즘으로 칼만 필터와 광류를 이용하는 영상기반 자세추정 알고리즘을 제안한다. 광류는 UAV에 장착된 카메라의 영상으로부터 획득하며 UAV의 자세는 광류를 통해 측정된다. 이 논문에서 UAV 자세의 추정과 낮은 신뢰성을 보완하기위해 칼만 필터를 사용한다. 그리고 실험을 통해 알고리즘을 검증하였다. Rate table과 실제 비행영상을 이용하여 실험 하였으며, 본 논문에서 UAV의 자세 추정 알고리즘 검증 결과를 보였다. Rate table 실험에서 오차는 2도 이내였으며, AHRS를 통해 측정한 결과와 비슷한 경향을 보인다. 그러나 실제 비행 영상 실험에서 최대 Yaw 오차는 21도였으며, 최대 Pitch 오차는 7.8도로 나타났다.

지진을 받는 다자유도 건물의 μ합성 제어기 설계 및 검증실험 (μ-Synthesis Controller Design and Experimental Verification for a Seismic-excited MDOF Building)

  • 민경원;주석준;이영철
    • 한국지진공학회논문집
    • /
    • 제6권6호
    • /
    • pp.41-48
    • /
    • 2002
  • 이 논문은 지진을 받는 축소 3층 건물모델의 제어 실험에 관한 것이다. 건물모델에 대해 능동 질량 감쇠기를 사용하여 강인제어기인 $\mu$합성 제어기를 설계하고 실험적으로 검증하였다. 외란 입력과 제어용 출력에 적합한 주파수 가중함수를 설정하고 가중함수의 조합을 분석하여 최적의 $\mu$합성 제어기를 제시하였다. 실험 결과 백색잡음 가진 입력에 대한 RMS 응답에서 60-70%의 감소를 보여주었고, 축소 지진 입력에 대한 최대 응답에서 30-45% 정도의 감소효과를 나타내었다. 수치해석결과와 실험의 결과가 매우 정확히 일치하였으며 불확실성을 갖는 구조물에 대하여 제어기가 불확실성의 범위 내에서 강인하다는 것을 보여주었다.

하수 내 총질소 성분 제거를 위한 YPNR 공정의 실증 연구 (Practical Demonstration of YPNR Process to Elimination the Total Nitrogen Ingredient in Sewage)

  • 임은태;정귀택;방성훈;김용운;박재희;박석환;박돈희
    • KSBB Journal
    • /
    • 제24권3호
    • /
    • pp.291-295
    • /
    • 2009
  • 본 연구는 오 폐수 내에 존재하는 총질소 성분의 생물학적 처리공정의 실증께 관한 연구를 수행한 것으로써 M면의 마을하수처리장에 YPNR 고도처리 공법을 적용하여 25일의 처리기간 동안 마을하수 중의 총질소 성분의 제거에 관한 자료를 확보하여 얻은 결론은 다음과 같다. 마을하수처리현장에 적용한 결과, 첨가탄소원 없이도 현재 배출허용기준에 적합한 70% 정도의 질소처리 효율을 나타내었다. 특히 유입 총질소 성분 중 95%는 질산화공정을 통해 질산성 질소로 전환되어 탈질공정에서 제거되고 방류되는 총질소 성분 중 질산성 질소 성분이 약 77%로서 이는 향후 강화되는 총질소 규제시 외부탄소원을 소량 투여해 줌으로써 현재의 방류수질보다 약 70% 낮은 방류수질을 유지 할 수 있을 것으로 생각된다. 결론적으로 본 실험에서 사용된 YPNR 고도처리 공정은 하수, 마을하수처리에 적용될 수 있다고 판단된다.

PSALI 연구를 위한 실물대 실험 장치와 축소 모형간의 유의성 검증 (Verification of Significancebetween Experiment Devices and Scaled-down Model for the Study of PSALI)

  • 이진숙;김소연;하태현;정용규
    • 조명전기설비학회논문지
    • /
    • 제25권12호
    • /
    • pp.11-20
    • /
    • 2011
  • PSALI is referred to the supplementary lighting for the interior lighting under the daily lighting situation, and pursuant to the pertinent regulations in energy savings design standard and others in recent architecture works, the importance thereof has been increasing gradually coupled with the energy performance index (EPI), energy savings plan and the like as well as expansion of submittal and implementation policies. However, this type of PSALI studies indeed have a number of limitations since it has surrounding environmental conditions in direction, season, region, climate, time, opening rate, window area ratio, actual index, reflection rate of finishing materials and others in the architecture work as well as frequent changes in interior lighting environment for variables in daily light volume flowing into the interior, and others. Therefore, this study has analyzed existing advance research cases to produce the actual-sized model and scaled-down model, and installed the artificial lighting of LED light source possible to reproduce with same capability on both models. As a result of comparison and analysis of the artificial lighting with the key light, it has certain level of error rate from the scaled down lighting device in certain rate and actual model butit was noticeably significant within specific scope.

연료분사 노즐 형상이 선박용 중형 디젤 엔진의 NOx에 미치는 영향 연구 (Effect of Fuel Nozzle Configuration on the Reduction of NOx Emission in Medium-speed Marine Diesel Engine)

  • 윤욱현;김병석;류승협;김기두;하지수
    • 한국마린엔지니어링학회:학술대회논문집
    • /
    • 한국마린엔지니어링학회 2005년도 후기학술대회논문집
    • /
    • pp.13-14
    • /
    • 2005
  • Multi-dimensional combustion analysis and experiment has been carried out to investigate the effects of the injector nozzle hole diameter and number on the NOx formation and fuel consumption in HYUNDAI HiMSEN engine. The behavior of spray and combustion phenomena in diesel engine was examined by FIRE code. Wave breakup and Zeldovich models were adopted to describe the atomization characteristics and NOx formation. Wallfilm model suggested by Mundo, et al. and auto-ignition model suggested by Theobald and Cheng were adopted to investigate the spray-wall interaction characteristics and ignition delay. The information of spray angle and spray tip penetration length was extracted from fuel spray visualization experiment and the fuel injection rate profile was extracted from fuel injection system experiment as an input and verification data for the combustion analysis. Next, the nine different nozzle configurations were simulated to evaluate the effect of injector hole diameter and number on the NOx formation and fuel consumption.

  • PDF

개인화 알고리즘으로 필터 버블이 형성되는 과정에 대한 검증 (A Verification about the Formation Process of Filter Bubble with Personalization Algorithm)

  • 전준영;황소윤;윤영미
    • 한국멀티미디어학회논문지
    • /
    • 제21권3호
    • /
    • pp.369-381
    • /
    • 2018
  • Nowadays a personalization algorithm is gaining huge attention. It gives users selective information which is helpful and interesting in a deluge of information based on their past behavior on the internet. However there is also a fatal side effect that the user can only get restricted information on restricted topics selected by the algorithm. Basically, the personalization algorithm makes users have a narrower perspective and even stronger bias because users have less chances to get views of opponent. Eli Pariser called this problem the 'filter bubble' in his book. It is important to understand exactly what a filter bubble is to solve the problem. Therefore, this paper shows how much Google's personalized search algorithm influences search result through an experiment with deep neural networks acting like users. At the beginning of the experiment, two Google accounts are newly created, not to be influenced by the Google's personalized search algorithm. Then the two pure accounts get politically biased by two methods. We periodically calculate the numerical score depending on the character of links and it shows how biased the account is. In conclusion, this paper shows the formation process of filter bubble by a personalization algorithm through the experiment.