• Title/Summary/Keyword: parameter function

Search Result 2,957, Processing Time 0.028 seconds

Performance analysis of weakly-supervised sound event detection system based on the mean-teacher convolutional recurrent neural network model (평균-교사 합성곱 순환 신경망 모델을 이용한 약지도 음향 이벤트 검출 시스템의 성능 분석)

  • Lee, Seokjin
    • The Journal of the Acoustical Society of Korea
    • /
    • v.40 no.2
    • /
    • pp.139-147
    • /
    • 2021
  • This paper introduces and implements a Sound Event Detection (SED) system based on weakly-supervised learning where only part of the data is labeled, and analyzes the effect of parameters. The SED system estimates the classes and onset/offset times of events in the acoustic signal. In order to train the model, all information on the event class and onset/offset times must be provided. Unfortunately, the onset/offset times are hard to be labeled exactly. Therefore, in the weakly-supervised task, the SED model is trained by "strongly labeled data" including the event class and activations, "weakly labeled data" including the event class, and "unlabeled data" without any label. Recently, the SED systems using the mean-teacher model are widely used for the task with several parameters. These parameters should be chosen carefully because they may affect the performance. In this paper, performance analysis was performed on parameters, such as the feature, moving average parameter, weight of the consistency cost function, ramp-up length, and maximum learning rate, using the data of DCASE 2020 Task 4. Effects and the optimal values of the parameters were discussed.

Complementary measures for Environmental Performance Evaluation Index of External Space of Green Standard for Energy and Environmental Design for Apartment Complex - Focused on the Respect of Response to Climate Change - (공동주택 녹색건축인증기준의 외부공간 환경성능 평가지표 보완방안 - 기후변화 대응 측면을 중심으로 -)

  • Ye, Tae-Gon;Kim, Kwang-Hyun;Kwon, Young-Sang
    • Journal of the Architectural Institute of Korea Planning & Design
    • /
    • v.34 no.1
    • /
    • pp.3-14
    • /
    • 2018
  • An apartment complex is a building use with great potential to contribute to solving problems related to urban ecological environment and climate change. The first goal of this study is to grasp the current situation of application and limitations of the ecological area rate, which is a representative evaluation index used to evaluate the environmental performance of the external space of an apartment complex in Green Standard for Energy and Environmental Design (G-SEED). The second goal is to propose a prototype of the evaluation index for evaluating greenhouse gas (GHG) reduction performance in order to supplement the evaluation index for the environmental performance of the external space in terms of response to climate change. We analyzed 43 cases of apartment complexes certified according to G-SEED, which was enforced since July 1, 2010, and found application characteristics of each space type and the limitations of ecological area rate. We analyzed overseas green building certification systems such as LEED and BREEAM that derived implications for supplementing the limitations of ecological area rate, which is focused on the evaluation of soil and water circulation function, and set up a development direction of complementary measures. Through analysis of previous studies, relevant regulations and standards, and technical documents of the manufacturer, the heat island mitigation performance of the pavement and roof surfaces of the apartment complex and the carbon uptake performance of the trees in the apartment complex was selected as parameters to yield the GHG reduction performance of the external space of the apartment complex. Finally, a quantitative evaluation method for each parameter and a prototype of the evaluation index for the GHG reduction performance were proposed. As a result of applying the prototype to an apartment complex case, the possibility of adoption and applicability as an evaluation index of G-SEED were proved.

Design of Processor Lever Controller for Electric Propulsion System of Naval Ship (전기추진 함정용 프로세서 레버 제어기 설계)

  • Shim, Jaesoon;Lee, Hunseok;Jung, Sung-Young;Oh, Jin-Seok
    • Journal of the Korea Institute of Information and Communication Engineering
    • /
    • v.25 no.1
    • /
    • pp.134-145
    • /
    • 2021
  • It is common to optimize the propulsion control system through a so-called tuning process that modifies the parameter values of the propulsion control software during a ship commissioning. However, during this process, if the error of the initial setting value is large, the tuning time may take too long, or the propulsion equipment can be seriously damaged. Therefore, we conducted research on the design of a propulsion controller that applied a Processor lever controller even for inexperienced people with relatively little experience in tuning propulsion control software to be able to reduce the tuning time while protecting the propulsion system. Through simulation, by comparing the execution result of propulsion control lever commands through the PI controller without applying the Processor lever controller. We analyzed the improvement of the Overshoot and propulsion performance. The simulation results showed that the safety of the propulsion system increased because Overshoot of approximately 9.74%, which occurred when the Processor lever function was not applied.

Drought evaluation using unstructured data: a case study for Boryeong area (비정형 데이터를 활용한 가뭄평가 - 보령지역을 중심으로 -)

  • Jung, Jinhong;Park, Dong-Hyeok;Ahn, Jaehyun
    • Journal of Korea Water Resources Association
    • /
    • v.53 no.12
    • /
    • pp.1203-1210
    • /
    • 2020
  • Drought is caused by a combination of various hydrological or meteorological factor, so it is difficult to accurately assess drought event, but various drought indices have been developed to interpret them quantitatively. However, the drought indexes currently being used are calculated from the lack of a single variable, which is a problem that does not accurately determine the drought event caused by complex causes. Shortage of a single variable may not be a drought, but it is judged to be a drought. On the other hand, research on developing indices using unstructured data, which is widely used in big data analysis, is being carried out in other fields and proven to be superior. Therefore, in this study, we intend to calculate the drought index by combining unstructured data (news data) with weather and hydrologic information (rainfall and dam inflow) that are being used for the existing drought index, and to evaluate the utilization of drought interpretation through verification of the calculated drought index. The Clayton Copula function was used to calculate the joint drought index, and the parameter estimation was used by the calibration method. The analysis showed that the drought index, which combines unstructured data, properly expresses the drought period compared to the existing drought index (SPI, SDI). In addition, ROC scores were calculated higher than existing drought indices, making them more useful in drought interpretation. The joint drought index calculated in this study is considered highly useful in that it complements the analytical limits of the existing single variable drought index and provides excellent utilization of the drought index using unstructured data.

Review on factors affecting the optical properties of dental zirconia (치과용 지르코니아의 광학적 성질에 영향을 미치는 요소에 대한 문헌고찰)

  • Park, Chan-Ho;Ko, Kyung-Ho;Park, Chan-Jin;Cho, Lee-Ra;Huh, Yoon-Hyuk
    • Journal of Dental Rehabilitation and Applied Science
    • /
    • v.37 no.4
    • /
    • pp.177-185
    • /
    • 2021
  • Clinical applications of translucent zirconia as well as traditional zirconia (3 mol% yttria stabilized tetragonal zirconia polycrystal, 3Y-TZP) are increasing. For this reason, studies on factors affecting the optical properties of dental zirconia have been continuously reported. The optical effect of dental zirconia may vary depending on the yttria content, the thickness of the prosthesis, the sintering process, polishing, glazing and cementation in laboratory and clinical procedures. Increasing the yttria concentration can reduce the masking effect. Translucency decreases as the thickness of the restoration increases, but the required thickness may vary depending on the properties of the zirconia block. The high-speed sintering method can shorten the manufacturing time, but in some cases, the translucency of the prosthesis may decrease. In addition, the optical properties can be affected by the surface roughness of zirconia and the polishing process. The use of an appropriate colored cement can help with the masking effect of zirconia and can be useful for color matching for more esthetic results.

A ScanSAR Processing without Azimuth Stitching by Time-domain Cross-correlation (Azimuth Stitching 없는 ScanSAR 영상화: 시간영역 교차상관)

  • Won, Joong-Sun
    • Korean Journal of Remote Sensing
    • /
    • v.38 no.3
    • /
    • pp.251-263
    • /
    • 2022
  • This paper presents an idea of ScanSAR image formation. For image formation of ScanSAR that utilizes the burst mode for raw signal acquisition, most conventional single burst methods essentially require a step of azimuth stitching which contributes to radiometric and phase distortions to some extent. Time-domain cross correlation could replace SPECAN which is most popularly used for ScanSAR processing. The core idea of the proposed method is that it is possible to relieve the necessity of azimuth stitching by an extension of Doppler bandwidth of the reference function to the burst cycle period. Performance of the proposed method was evaluated by applying it to the raw signals acquired by a spaceborne SAR system, and results satisfied all image quality requirements including 3 dB width, peak-to-sidelobe ratio (PSLR), compression ratio,speckle noise, etc. Image quality of ScanSAR is inferior to that of Stripmap in all aspects. However, it is also possible to improve the quality of ScanSAR image competitive to that of Stripmap if focused on a certain parameter while reduced qualities of other parameters. Thus, it is necessary for a ScanSAR processor to offer a great degree of flexibility complying with different requirements for different applications and techniques.

Regionalization of rainfall-runoff model parameters based on the correlation of regional characteristic factors (지역특성인자의 상호연관성을 고려한 강우-유출모형 매개변수 지역화)

  • Kim, Jin-Guk;Sumyia, Uranchimeg;Kim, Tae-Jeong;Kwon, Hyun-Han
    • Journal of Korea Water Resources Association
    • /
    • v.54 no.11
    • /
    • pp.955-968
    • /
    • 2021
  • A water resource plan is routinely based on a natural flow and can be estimated using observed streamflow data or a long-term continuous rainfall-runoff model. However, the watershed with the natural flow is very limited to the upstream area of the dam. In particular, for the ungauged watershed, a rainfall-runoff model is established for the gauged watershed, and the model is then applied to the ungauged watershed by transferring the associated parameters. In this study, the GR4J rainfall-runoff model is mainly used to regionalize the parameters that are estimated from the 14 dam watershed via an optimization process. In terms of optimizing the parameters, the Bayesian approach was applied to consider the uncertainty of parameters quantitatively, and a number of parameter samples obtained from the posterior distribution were used for the regionalization. Here, the relationship between the estimated parameters and the topographical factors was first identified, and the dependencies between them are effectively modeled by a Copula function approach to obtain the regionalized parameters. The predicted streamflow with the use of regionalized parameters showed a good agreement with that of the observed with a correlation of about 0.8. It was found that the proposed regionalized framework is able to effectively simulate streamflow for the ungauged watersheds by the use of the regionalized parameters, along with the associated uncertainty, informed by the basin characteristics.

Measurements of Void Concentration Parameters in the Drift-Flux Model (상대유량 모델내의 기포분포계수 측정에 관한 연구)

  • Yun, B.J.;Park, G.C.;Chung, C.H.
    • Nuclear Engineering and Technology
    • /
    • v.25 no.1
    • /
    • pp.91-101
    • /
    • 1993
  • To predict accurately the thermal hydraulic behavior of light water reactors during normal or abnormal operation, the accurate estimation of the void distribution is required. Up to date, many techniques for predicting void fraction of two-phase flow systems have been suggested. Among these techniques, the drift-flux model is widely used because of its exact calculation ability and simplicity. However, to get more accurate prediction of void fraction using drift-flux model, slip and flow regime effects must be considered more properly In the drift-flux method, these two effects are accounted for by two drift-flux parameters ; $C_{o}$ and (equation omitted). At earlier stage, $C_{o}$ is measured in a circular tube. In this study, $C_{o}$ is experimentally determined by measuring local void fraction and vapor velocity distribution in a rectangular subchannel having 4 heating rods which simulates nuclear subchannels. The measurements are peformed with two-electrical conductivity probes which are known to be adequate for measuring local parameters. The experiments are performed at low flow rate and the system pressure less than 3 atmo spheric pressure. In this experiment, (equation omitted), is not measured, but quoted from well-known empirical correlation to formulate $C_{o}$. Finally, $C_{o}$ is expressed as a function of channel averaged void fraction. fraction.

  • PDF

Multi-camera Calibration Method for Optical Motion Capture System (광학식 모션캡처를 위한 다중 카메라 보정 방법)

  • Shin, Ki-Young;Mun, Joung-H.
    • Journal of the Korea Society of Computer and Information
    • /
    • v.14 no.6
    • /
    • pp.41-49
    • /
    • 2009
  • In this paper, the multi-camera calibration algorithm for optical motion capture system is proposed. This algorithm performs 1st camera calibration using DLT(Direct linear transformation} method and 3-axis calibration frame with 7 optical markers. And 2nd calibration is performed by waving with a wand of known length(so called wand dance} throughout desired calibration volume. In the 1st camera calibration, it is obtained not only camera parameter but also radial lens distortion parameters. These parameters are used initial solution for optimization in the 2nd camera calibration. In the 2nd camera calibration, the optimization is performed. The objective function is to minimize the difference of distance between real markers and reconstructed markers. For verification of the proposed algorithm, re-projection errors are calculated and the distance among markers in the 3-axis frame and in the wand calculated. And then it compares the proposed algorithm with commercial motion capture system. In the 3D reconstruction error of 3-axis frame, average error presents 1.7042mm(commercial system) and 0.8765mm(proposed algorithm). Average error reduces to 51.4 percent in commercial system. In the distance between markers in the wand, the average error shows 1.8897mm in the commercial system and 2.0183mm in the proposed algorithm.

Feeding by common heterotrophic protists on the mixotrophic dinoflagellate Ansanella granifera (Suessiaceae, Dinophyceae)

  • Hee Chang Kang;Hae Jin Jeong;An Suk Lim;Jin Hee Ok;Ji Hyun You;Sang Ah Park;Se Hee Eom
    • ALGAE
    • /
    • v.38 no.1
    • /
    • pp.57-70
    • /
    • 2023
  • The mortality rate of red-tide dinoflagellates owing to predation is a major parameter that affects their population dynamics. The dinoflagellates Ansanella granifera and Ansanella sp. occasionally cause red tides. To understand the interactions between common heterotrophic protists and A. granifera, we explored the feeding occurrence of nine heterotrophic protists on A. granifera and the growth and ingestion rates of the heterotrophic dinoflagellate Gyrodinium dominans on A. granifera as a function of prey concentration and those of Oxyrrhis marina at a single high prey concentration. The heterotrophic dinoflagellates Aduncodinium glandula, G. dominans, Gyrodinium moestrupii, Luciella masanensis, Oblea rotunda, O. marina, Polykrikos kofoidii, and Pfiesteria piscicida and the naked ciliate Strombidium sp. were able to feed on A. granifera. With increasing mean prey concentrations, the growth and ingestion rates of G. dominans feeding on A. granifera rapidly increased and became saturated or slowly increased. The maximum growth and ingestion rates of G. dominans on A. granifera were 0.305 d-1 and 0.42 ng C predator-1 d-1 (3.8 cells predator-1 d-1), respectively. Furthermore, the growth and ingestion rates of O. marina on A. granifera at 1,700 ng C mL-1 (15,454 cells mL-1) were 0.037 d-1 and 0.19 ng C predator-1 d-1 (1.7 cells predator-1 d-1), respectively. The growth and ingestion rates of G. dominans and O. marina feeding on A. granifera were almost the lowest among those on the dinoflagellate prey species. Therefore, G. dominans and O. marina may prefer A. granifera less than other dinoflagellate prey species. The low mortality rate of A. granifera may positively affect its bloom formation.