• 제목/요약/키워드: multi-scale technique

검색결과 202건 처리시간 0.027초

Scaling Up Face Masks Classification Using a Deep Neural Network and Classical Method Inspired Hybrid Technique

  • Kumar, Akhil;Kalia, Arvind;Verma, Kinshuk;Sharma, Akashdeep;Kaushal, Manisha;Kalia, Aayushi
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • 제16권11호
    • /
    • pp.3658-3679
    • /
    • 2022
  • Classification of persons wearing and not wearing face masks in images has emerged as a new computer vision problem during the COVID-19 pandemic. In order to address this problem and scale up the research in this domain, in this paper a hybrid technique by employing ResNet-101 and multi-layer perceptron (MLP) classifier has been proposed. The proposed technique is tested and validated on a self-created face masks classification dataset and a standard dataset. On self-created dataset, the proposed technique achieved a classification accuracy of 97.3%. To embrace the proposed technique, six other state-of-the-art CNN feature extractors with six other classical machine learning classifiers have been tested and compared with the proposed technique. The proposed technique achieved better classification accuracy and 1-6% higher precision, recall, and F1 score as compared to other tested deep feature extractors and machine learning classifiers.

An Approach to Improve the Contrast of Multi Scale Fusion Methods

  • Hwang, Tae Hun;Kim, Jin Heon
    • Journal of Multimedia Information System
    • /
    • 제5권2호
    • /
    • pp.87-90
    • /
    • 2018
  • Various approaches have been proposed to convert low dynamic range (LDR) to high dynamic range (HDR). Of these approaches, the Multi Scale Fusion (MSF) algorithm based on Laplacian pyramid decomposition is used in many applications and demonstrates its usefulness. However, the pyramid fusion technique has no means for controlling the luminance component because the total number of pixels decreases as the pyramid rises to the upper layer. In this paper, we extract the reflection light of the image based on the Retinex theory and generate the weight map by adjusting the reflection component. This weighting map is applied to achieve an MSF-like effect during image fusion and provides an opportunity to control the brightness components. Experimental results show that the proposed method maintains the total number of pixels and exhibits similar effects to the conventional method.

A Multi-Level Simulation Technique for Large-ScaleAnalog Integrated Circuits

  • Yang Jeemo
    • 한국산업정보학회:학술대회논문집
    • /
    • 한국산업정보학회 1998년도 공동추계학술대회 경제위기 극복을 위한 정보기술의 효율적 활용
    • /
    • pp.827-834
    • /
    • 1998
  • This paper describes a multi-level simulation technique and its implementation, which accurately solve voltages and currents of circuits descreibed at mixed levels of abstractions. A metho to form a tightly coupled simulation environment is proposed and, starting from a description of a circuit, simulation set-up and analysis procedure of the multi-level simulator for a transient response are presented. Circuit and behavioral simulation techniques and their implementations composing the multi-level simulation are explained in detail. Most of the algorithms implemented in the simulation are based upon the standard simulation techniques in order to obtain the reliability and accuracy of conventinoal simulators. Simulation examples show that the multi-level simulator can analyze circuits containing highly nonlinear behavioral models without loss of accuracy provided the behavioral models are accurate enough.

확장유한요소법과 멀티스케일 기법을 통한 팔라듐 첨가 탄소섬유/알루미늄 적층구조에 대한 수치해석 (Numerical Analysis of Palladium added Carbon Fiber/Al using Extended Finite Element Method and Multiscale Technique)

  • 박우림;권오헌
    • 한국안전학회지
    • /
    • 제34권2호
    • /
    • pp.7-14
    • /
    • 2019
  • A palladium can adsorb hydrogen and detect leaking hydrogen through changes in color and electrical resistance. This study is to evaluate the structural behavior of carbon fiber adding palladium composite materials used in the hydrogen storage vessel. A multi-scale analysis technique was used to analyze accurately the behavior of each material in relation to the microscopic composition. The multi-scale analysis is more proper and precise for composite materials because of considering the individual microscopic structure and properties of each material for composite materials. Also the crack evaluation was performed by XFEM analysis to confirm the reinforcement performance of aluminum as a liner of the hydrogen vessel. The results show that the addition of the palladium material increased the macroscopic stress, but microscopically the carbon fiber stress was reduced. It means the performance improvement of the palladium added carbon fiber/Al composite.

Implementation Strategy for the Numerical Efficiency Improvement of the Multiscale Interpolation Wavelet-Galerkin Method

  • Seo Jeong Hun;Earmme Taemin;Jang Gang-Won;Kim Yoon Young
    • Journal of Mechanical Science and Technology
    • /
    • 제20권1호
    • /
    • pp.110-124
    • /
    • 2006
  • The multi scale wavelet-Galerkin method implemented in an adaptive manner has an advantage of obtaining accurate solutions with a substantially reduced number of interpolation points. The method is becoming popular, but its numerical efficiency still needs improvement. The objectives of this investigation are to present a new numerical scheme to improve the performance of the multi scale adaptive wavelet-Galerkin method and to give detailed implementation procedure. Specifically, the subdomain technique suitable for multiscale methods is developed and implemented. When the standard wavelet-Galerkin method is implemented without domain subdivision, the interaction between very long scale wavelets and very short scale wavelets leads to a poorly-sparse system matrix, which considerably worsens numerical efficiency for large-sized problems. The performance of the developed strategy is checked in terms of numerical costs such as the CPU time and memory size. Since the detailed implementation procedure including preprocessing and stiffness matrix construction is given, researchers having experiences in standard finite element implementation may be able to extend the multi scale method further or utilize some features of the multiscale method in their own applications.

Multi-Domain과 행렬 보간법을 이용한 강 하구와 연안의 2차원 수치해석 (Numerical Simulation of 2-D Estuaries and Coast by Multi-Domain and the Interpolating Matrix Method)

  • 채현식
    • 한국전산유체공학회지
    • /
    • 제2권1호
    • /
    • pp.21-28
    • /
    • 1997
  • This paper presents a two-dimensional horizontal implicit model to general circulation in estuaries and coastal seas. The model is developed in non-orthogonal curvilinear coordinates system, using the Interpolating Matrix Method (IMM), in combination with a technique of multi-domain. In the propose model, the Saint-Venant equations are solved by a splitting-up technique, in the successive steps; convection, diffusion and wave propagation. The ability of the proposed model to deal with full scale nature is illustrated by the interpretation of a dye-tracing experiment in the Gironde estuary.

  • PDF

Stochastic upscaling via linear Bayesian updating

  • Sarfaraz, Sadiq M.;Rosic, Bojana V.;Matthies, Hermann G.;Ibrahimbegovic, Adnan
    • Coupled systems mechanics
    • /
    • 제7권2호
    • /
    • pp.211-232
    • /
    • 2018
  • In this work we present an upscaling technique for multi-scale computations based on a stochastic model calibration technique. We consider a coarse-scale continuum material model described in the framework of generalized standard materials. The model parameters are considered uncertain, and are determined in a Bayesian framework for the given fine scale data in a form of stored energy and dissipation potential. The proposed stochastic upscaling approach is independent w.r.t. the choice of models on coarse and fine scales. Simple numerical examples are shown to demonstrate the ability of the proposed approach to calibrate coarse scale elastic and inelastic material parameters.

SOM의 통계적 특성과 다중 스케일 Bayesian 영상 분할 기법을 이용한 텍스쳐 분할 (Texture Segmentation Using Statistical Characteristics of SOM and Multiscale Bayesian Image Segmentation Technique)

  • 김태형;엄일규;김유신
    • 대한전자공학회논문지SP
    • /
    • 제42권6호
    • /
    • pp.43-54
    • /
    • 2005
  • 이본 논문에서는 Bayesian 영상 분할법과 SOM(Self Organization feature Map)을 이용한 텍스쳐(Texture) 분할 방법을 제안한다. SOM의 입력으로 다중 스케일에서의 웨이블릿 계수를 사용하고, 훈련된 SOM으로부터 관측 데이터에 대한 우도(尤度, likelihood)와 사후확률을 구하는 방법을 제시한다. 훈련된 SOM들로부터 구한 사후확률과 MAP(Maximum A Posterior) 분류법을 이용하여 텍스쳐 분할을 얻는다. 그리고 문맥 정보를 이용하여 텍스쳐 분할 결과를 개선하였다. 제안 방법은 HMT(Hidden Markov Tree)을 이용한 텍스쳐 분할보다 더 우수한 결과를 보여준다. 또한 SOM과 HMTseg라고 불리는 다중스케일 Bayesian 영상 분할 기법을 이용한 텍스쳐 분할 결과는 HMT와 HMTseg을 이용한 결과보다 더 우수한 성능을 보여준다.

절점 비활성화 기법을 적용한 나노-연속체 멀티스케일 해석 기법 (Nano-continuum multi scale analysis using node deactivation techniques)

  • 이승윤;조맹효
    • 한국전산구조공학회:학술대회논문집
    • /
    • 한국전산구조공학회 2006년도 정기 학술대회 논문집
    • /
    • pp.395-402
    • /
    • 2006
  • In analyzing the nano-scale phenomena or behaviors of nano devices or materials, it is often desirable to deal with more atoms than can be treated only with a full atomistic simulation. However, even now, it is advisable to apply the atomistic simulation to the narrow region where the deformation field changes rapidly but to apply the conventional continuum model to the region far from that region. This equivalent continuum model can be formulated by applying the Cauchy-Born rule to the exact atomistic potential as in the quasicontinuum method. To couple the atomistic model with the equivalent continuum model, continuum displacements are conformed to the molecular displacements at the discrete positions of the atoms within the bridging domain. To satisfy the coupling constraints, we apply the Lagrange multiplier method. The continuum model in the bridging model should be applied on the region where the deformation field changes gradually. Then we can make the nodal spacing in the continuum model be much larger than the atomic spacing. In the first step, we generate the atomic-resolution mesh with the nodal spacing equal to the atomic spacing, and then we eliminate the nodal degrees of freedom adaptively using the node deactivation techniques. We eliminate more DOFs as the regions are more far from the atomistic region. Computing time and computational resources can be greatly reduced by the present node deactivation technique in multi scale analysis.

  • PDF

기름방울 형상 및 그 체적 분석법 (Droplet Geometry and Its Volume Analysis)

  • 윤문철
    • Tribology and Lubricants
    • /
    • 제24권6호
    • /
    • pp.320-325
    • /
    • 2008
  • The recent industrial application requires technical methods to get the cutting fluid droplet surfaces in particular from the viewpoint of topography and micro texture. To characterize the surface topography of droplet, the combination of the confocal laser scanning microscope (CLSM) and wavelet filtering is well suited for obtaining the droplet geometry encountered in tribological research. This technique indicates a better agreement in obtaining an appropriate droplet surface obtained by the CLSM over a detail range of surface accuracy (resolution: $2{\mu}m$). And the results allow an excellent accuracy in a measurement of a droplet surface. The combination of extended focal depth measurement configured and multi-scale wavelet filtering has proven that it can construct a droplet surface in a successive and accurate way. A multi-scale approach of wavelet filtering was developed based on the decomposition and reconstruction of droplet surface by 2D wavelet transform using db9 (a mother wavelet of daubechies). Also this technique can be extended to characterize the quantification of droplet properties and other field in a wide range of scales. Finally this method is verified to be a better droplet surface modeling in a micro scale arising in a mist machining.