• 제목/요약/키워드: Test Network

검색결과 3,532건 처리시간 0.026초

Neural Network Image Reconstruction for Magnetic Particle Imaging

  • Chae, Byung Gyu
    • ETRI Journal
    • /
    • 제39권6호
    • /
    • pp.841-850
    • /
    • 2017
  • We investigate neural network image reconstruction for magnetic particle imaging. The network performance strongly depends on the convolution effects of the spectrum input data. The larger convolution effect appearing at a relatively smaller nanoparticle size obstructs the network training. The trained single-layer network reveals the weighting matrix consisting of a basis vector in the form of Chebyshev polynomials of the second kind. The weighting matrix corresponds to an inverse system matrix, where an incoherency of basis vectors due to low convolution effects, as well as a nonlinear activation function, plays a key role in retrieving the matrix elements. Test images are well reconstructed through trained networks having an inverse kernel matrix. We also confirm that a multi-layer network with one hidden layer improves the performance. Based on the results, a neural network architecture overcoming the low incoherence of the inverse kernel through the classification property is expected to become a better tool for image reconstruction.

Video Quality Assessment based on Deep Neural Network

  • Zhiming Shi
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • 제17권8호
    • /
    • pp.2053-2067
    • /
    • 2023
  • This paper proposes two video quality assessment methods based on deep neural network. (i)The first method uses the IQF-CNN (convolution neural network based on image quality features) to build image quality assessment method. The LIVE image database is used to test this method, the experiment show that it is effective. Therefore, this method is extended to the video quality assessment. At first every image frame of video is predicted, next the relationship between different image frames are analyzed by the hysteresis function and different window function to improve the accuracy of video quality assessment. (ii)The second method proposes a video quality assessment method based on convolution neural network (CNN) and gated circular unit network (GRU). First, the spatial features of video frames are extracted using CNN network, next the temporal features of the video frame using GRU network. Finally the extracted temporal and spatial features are analyzed by full connection layer of CNN network to obtain the video quality assessment score. All the above proposed methods are verified on the video databases, and compared with other methods.

결정론적 테스트 세트의 신호확률에 기반을 둔 clustered reconfigurable interconnection network 내장된 자체 테스트 기법 (A Clustered Reconfigurable Interconnection Network BIST Based on Signal Probabilities of Deterministic Test Sets)

  • 송동섭;강성호
    • 대한전자공학회논문지SD
    • /
    • 제42권12호
    • /
    • pp.79-90
    • /
    • 2005
  • 본 논문에서는 의사무작위패턴만으로는 생산하기 힘든 결정론적 테스트 큐브의 생산확률을 높일 수 있는 새로운 clustered reconfigurable interconnect network (CRIN) 내장된 자체 테스트 기법을 제안한다. 제안된 방법은 주어진 테스트 큐브들의 신호확률에 기반을 둔 스캔 셀 재배치 기술과 규정 비트(care-bit: 0 또는 1)가 집중된 스캔 체인 테스트 큐브의 생산확률을 높이기 위한 전용의 하드웨어 블록을 사용한다. 테스트 큐브의 생산확률을 최대로 할 수 있는 시뮬레이티드 어닐링(simulated annealing) 기반 알고리듬이 스캔 셀 재배치를 위해 개발되었으며, CRIN 하드웨어 합성을 위한 반복 알고리듬 또한 개발되었다. 실험을 통하여 제안된 CRIN 내장된 자체 테스트 기법은 기존의 연구 결과보다 훨씬 적은 저장 공간과 짧은 테스트 시간으로 $100\%$의 고장검출율을 달성할 수 있음을 증명한다.

정밀 GPS 해석 S/W GMAIT/GLOBK를 활용한 TBM의 3차원 위치 결정 (Determination of 3-D Positions on TBMs Using the Precise GPS Data analysis SW, GAMIT/GLOBK)

  • 유경완;양인태;이동하
    • 산업기술연구
    • /
    • 제36권
    • /
    • pp.71-76
    • /
    • 2016
  • In this study, we determined the precise coordinates of TBMs (Tidal Bench Marks), which used as the national reference points in coastal area of Korea, using a GPS data analysis SW for the academic and scientific applications, GAMIT/GLOBK. For accurate 3-D positioning of TBM locations, we performed the GPS point surveying according to the national surveying policy and also acquired the GPS data for 48 TBMs located in the western and southern coastal part of Korea. Considering the results of baseline analysis to each observation session obtained from GAMIT module, the baseline analysis was realized to be done precisely because the values of Normalized RMS (NRMS) were mostly less than ${\pm}0.20mm$. Before the network adjustment using GLOBK module, we evaluated the suitability of observations for each session by applying the chi-squared test (${\chi}^2$ test) to the degree of freedom in observed session. An overall distributions of ${\chi}^2$ test were less than 1.0 for all sessions, and the statistical of ${\chi}^2$ test showed the average, 0.267 with minimum and maximum value, 0.063 and 0.653, respectively. Finally, we analyzed the network adjustment for 48 TBMs to reduce the residuals of baseline analysis on each point by connecting with 42 permanent GPS stations in Korea. In the network adjustment procedure, we set up the weighted values of each permanent station to be allocated between 0.9 and 1.14, and also removed the observed points having residual exceeds 4-times of standard deviation ($4{\sigma}$).

  • PDF

A Novel Test Structure for Process Control Monitor for Un-Cooled Bolometer Area Array Detector Technology

  • Saxena, R.S.;Bhan, R.K.;Jalwania, C.R.;Lomash, S.K.
    • JSTS:Journal of Semiconductor Technology and Science
    • /
    • 제6권4호
    • /
    • pp.299-312
    • /
    • 2006
  • This paper presents the results of a novel test structure for process control monitor for uncooled IR detector technology of microbolometer arrays. The proposed test structure is based on resistive network configuration. The theoretical model for resistance of this network has been developed using 'Compensation' and 'Superposition' network theorems. The theoretical results of proposed resistive network have been verified by wired hardware testing as well as using an actual 16x16 networked bolometer array. The proposed structure uses simple two-level metal process and is easy to integrate with standard CMOS process line. The proposed structure can imitate the performance of actual fabricated version of area array closely and it uses only 32 pins instead of 512 using conventional method for a $16{\times}16$ array. Further, it has been demonstrated that the defective or faulty elements can be identified vividly using extraction matrix, whose values are quite similar(within the error of 0.1%), which verifies the algorithm in small variation case(${\sim}1%$ variation). For example, an element, intentionally damaged electrically, has been shown to have the difference magnitude much higher than rest of the elements(1.45 a.u. as compared to ${\sim}$ 0.25 a.u. of others), confirming that it is defective. Further, for the devices having non-uniformity ${\leq}$ 10%, both the actual non-uniformity and faults are predicted well. Finally, using our analysis, we have been able to grade(pass or fail) 60 actual devices based on quantitative estimation of non-uniformity ranging from < 5% to > 20%. Additionally, we have been able to identify the number of bad elements ranging from 0 to > 15 in above devices.

NGN 환경에 적합한 단대단 성능평가 방법 (The End-to-end Performance Evaluation Method on Next Generation Network)

  • 남창섭;김형수;김남;전홍범
    • 한국통신학회논문지
    • /
    • 제31권10B호
    • /
    • pp.857-865
    • /
    • 2006
  • NGN(Next Generation Network)이 약속하는 다양한 기능과 서비스가 제공될 수 있는가를 시험하고 평가하기 위해 평가방법이 필요하며 시험 대상 장치의 구성, 성능평가대상, 시험에 필요한 트래픽의 특성 등에 대한 개발이 요구되고 있다. 그간 IP패킷 성능평가방법으로 시뮬레이션(simulation)방법과 분석적 모델링(analytical modeling)방법 등 많은 연구가 수행되어 왔지만 아직까지 미흡한 실정이다. 본 논문은 NGN 통신망에 대하여 실제 운용환경과 유사한 조건을 만족시키고, 객관적이고 신뢰성 있는 성능 평가 방법을 도출하여 제시하였다. 이를 위해 시험용 트래픽(Test Traffic)과 백그라운드 트래픽(Background Traffic)을 개발하였고, IP 통신망과 동일한 장비로 구축된 시험망을 구성하여 NGN 서비스 특성을 고려한 대역폭, 패킷사이즈별로 성능평가를 수행하였다. 제안된 평가방법을 통해, 지연(Delay), 지연변이(Jitter) 손실(Loss)등의 성능평가결과를 기존의 평가방법의 결과와 비교하였을 때 제안된 평가방법이 단대단NGN 통신망의 성능평가방법으로 우수함을 입증하였다.

망분리 네트워크 상황에서 사이버보안 취약점 실시간 보안관제 평가모델 (Real-time security Monitroing assessment model for cybersecurity vulnera bilities in network separation situations)

  • 이동휘;김홍기
    • 융합보안논문지
    • /
    • 제21권1호
    • /
    • pp.45-53
    • /
    • 2021
  • 망분리 네트워크에서 보안관제를 할 경우 내부망 또는 위험도가 높은 구간에서는 평시 이상징후 탐지가 거의 이루어지지 않는다. 그렇기 때문에 보안 네트워크 구축 후 최적화 된 보안구조를 완성하기 위해서 망분리된 내부방에서의 최신 사이버 위협 이상징후를 평가할 수 있는 모델이 필요하다. 본 연구에서 일반 네트워크와 망분리 네트워크에서 발생하는 사이버 취약점과 악성코드를 데이터셋으로 발생시켜 평가하여, 망분리 내부망 사이버 공격에 위협 분석 및 최신 사이버 취약점을 대비 할 수 있게 하고, 특성에 맞는 사이버 보안 테스트 평가 체계를 구축하였다. 이를 실제 망분리 기관에 적용 가능한 평가모델을 설계 하고, 테스트 망을 각 상황별로 구축하여 실시간 보안관제 평가 모델을 적용하였다.

Social Network Effects on Travel Agency Employees' Occupational Outcomes: Innovation Behavior as a Mediator

  • Lee, Byeong-Cheol
    • 유통과학연구
    • /
    • 제15권6호
    • /
    • pp.13-24
    • /
    • 2017
  • Purpose - The current study aims to examine the effect of social network factors on travel agency employees' occupational outcomes such as job performance and job satisfaction through innovation behavior in a comprehensive model. Research design, data, and methodology - Based on a theory of social network, the concept of social network was assessed by three factors: a) network size, b) network range, and c) tie strength. To test the proposed hypotheses, structural equation modeling (SEM) was employed based on data from 197 travel agency employees in Korea. Result - The results showed that the associational activity of network size had a positive effect on innovation behavior, while the network range of network size had a significant negative effect on innovation behavior. Subsequently, innovation behavior positively influenced on job performance and job satisfaction, respectively. Conclusions - The results offer some insights into the extended model and have important managerial implications for Korean travel agencies. More specifically, considering diverse domains of social network and organizational research, this study advances critical utility of social network factors in a high facilitating level of innovation behavior, which can help travel agency employees promote their job performance and job satisfaction.

Artificial neural network model using ultrasonic test results to predict compressive stress in concrete

  • Ongpeng, Jason;Soberano, Marcus;Oreta, Andres;Hirose, Sohichi
    • Computers and Concrete
    • /
    • 제19권1호
    • /
    • pp.59-68
    • /
    • 2017
  • This study focused on modeling the behavior of the compressive stress using the average strain and ultrasonic test results in concrete. Feed-forward backpropagation artificial neural network (ANN) models were used to compare four types of concrete mixtures with varying water cement ratio (WC), ordinary concrete (ORC) and concrete with short steel fiber-reinforcement (FRC). Sixteen (16) $150mm{\times}150mm{\times}150mm$ concrete cubes were used; each contained eighteen (18) data sets. Ultrasonic test with pitch-catch configuration was conducted at each loading state to record linear and nonlinear test response with multiple step loads. Statistical Spearman's rank correlation was used to reduce the input parameters. Different types of concrete produced similar top five input parameters that had high correlation to compressive stress: average strain (${\varepsilon}$), fundamental harmonic amplitude (A1), $2^{nd}$ harmonic amplitude (A2), $3^{rd}$ harmonic amplitude (A3), and peak to peak amplitude (PPA). Twenty-eight ANN models were trained, validated and tested. A model was chosen for each WC with the highest Pearson correlation coefficient (R) in testing, and the soundness of the behavior for the input parameters in relation to the compressive stress. The ANN model showed increasing WC produced delayed response to stress at initial stages, abruptly responding after 40%. This was due to the presence of more voids for high water cement ratio that activated Contact Acoustic Nonlinearity (CAN) at the latter stage of the loading path. FRC showed slow response to stress than ORC, indicating the resistance of short steel fiber that delayed stress increase against the loading path.