• Structural Monitoring and Maintenance
• /
• v.9 no.3
• /
• pp.289-303
• /
• 2022

In this study, the impact of assigned pixel labels on the accuracy of crack image identification of steel structures is examined by using an atrous separable convolution neural network (ASCNN). Firstly, images containing fatigue cracks collected from steel structures are classified into four datasets by assigning different pixel labels based on image features. Secondly, the DeepLab v3+ algorithm is used to determine optimal parameters of the ASCNN model by maximizing the average mean-intersection-over-union (mIoU) metric of the datasets. Thirdly, the ASCNN model is trained for various image sizes and hyper-parameters, such as the learning rule, learning rate, and epoch. The optimal parameters of the ASCNN model are determined based on the average mIoU metric. Finally, the trained ASCNN model is evaluated by using 10% untrained images. The result shows that the ASCNN model can segment cracks and other objects in the captured images with an average mIoU of 0.716.

• Journal of Korea Port Economic Association
• /
• v.31 no.4
• /
• pp.189-207
• /
• 2015

The purpose of this study is to examine the relationship between the network characteristics of ports and their performance that is represented by port competitiveness for the port operators. The study employs Social Network Analysis (SNA) to evaluate network characteristics comprising four centrality indices. For this research, data from Containerization International Yearbooks for 2006-2011 is used to analyze the service networks of 20 major liner shipping companies. In SNA, nodes (vertices) in the network are the ports and links (edges) in the network are connections realized by vessel movements, such that the liner shipping network determines the port network. In addition, panel regression analysis has been employed to investigate the relationship between port network characteristics and their performance. The results suggest that the four centrality indices identify the roles of the world's major ports from 2006 to 2011 and that port performance is determined not only by macroeconomic variables and service capabilities but also by the eigenvector centrality of ports in networks.

• Journal of Information Processing Systems
• /
• v.14 no.6
• /
• pp.1464-1479
• /
• 2018

In face recognition, sometimes the number of available training samples for single category is insufficient. Therefore, the performances of models trained by convolutional neural network are not ideal. The small sample face recognition algorithm based on novel Siamese network is proposed in this paper, which doesn't need rich samples for training. The algorithm designs and realizes a new Siamese network model, SiameseFacel, which uses pairs of face images as inputs and maps them to target space so that the $L_2$ norm distance in target space can represent the semantic distance in input space. The mapping is represented by the neural network in supervised learning. Moreover, a more lightweight Siamese network model, SiameseFace2, is designed to reduce the network parameters without losing accuracy. We also present a new method to generate training data and expand the number of training samples for single category in AR and labeled faces in the wild (LFW) datasets, which improves the recognition accuracy of the models. Four loss functions are adopted to carry out experiments on AR and LFW datasets. The results show that the contrastive loss function combined with new Siamese network model in this paper can effectively improve the accuracy of face recognition.

• Computers and Concrete
• /
• v.8 no.3
• /
• pp.279-292
• /
• 2011

The main purpose of this study includes investigation of the rheological properties of fresh self consolidating cement paste containing chemical and mineral additives using Artificial Neural Network (ANN) model. In order to develop the model, 200 different mixes are cast in the laboratory as a part of an extensive experimental research program. The data used in the ANN model are arranged in a format of fourteen input parameters covering water-binder ratio, four different mineral additives (calcium carbonate, metakaolin, silica fume, and limestone), five different superplasticizers based on the poly carboxylate and naphthalene and four different Viscosity Modified Admixtures (VMAs). Two common output parameters including the mini slump value and flow cone time are chosen for measuring the rheological properties of fresh self consolidating cement paste. Having validated the model, the influence of effective parameters on the rheological properties of fresh self consolidating cement paste is investigated based on the ANN model outputs. The output results of the model are then compared with the results of previous studies performed by other researchers. Ultimately, the analysis of the model outputs determines the optimal percentage of additives which has a strong influence on the rheological properties of fresh self consolidating cement paste. The proposed ANN model shows that metakaolin and silica fume affect the rheological properties in the same manner. In addition, for providing the suitable rheological properties, the ANN model introduces the optimal percentage of metakaolin, silica fume, calcium carbonate and limestone as 15, 15, 20 and 20% by cement weight, respectively.

• International journal of advanced smart convergence
• /
• v.10 no.1
• /
• pp.1-11
• /
• 2021

Document page segmentation is an important step in building a quality optical character recognition module. The study examined already existing work on the topic of page segmentation and focused on the development of a segmentation model that has greater functional significance for application in an organization, as well as broad capabilities for managing the quality of the model. The main problems of document segmentation were highlighted, which include a complex background of intersecting objects. As classes for detection, not only classic text, table and figure were selected, but also additional types, such as signature, logo and table without borders (or with partially missing borders). This made it possible to pose a non-trivial task of detecting non-standard document elements. The authors compared existing neural network architectures for object detection based on published research data. The most suitable architecture was RetinaNet. To ensure the possibility of quality control of the model, a method based on neural network modeling using the RetinaNet architecture is proposed. During the study, several models were built, the quality of which was assessed on the test sample using the Mean average Precision metric. The best result among the constructed algorithms was shown by a model that includes four neural networks: the focus of the first neural network on detecting tables and tables without borders, the second - seals and signatures, the third - pictures and logos, and the fourth - text. As a result of the analysis, it was revealed that the approach based on four neural networks showed the best results in accordance with the objectives of the study on the test sample in the context of most classes of detection. The method proposed in the article can be used to recognize other objects. A promising direction in which the analysis can be continued is the segmentation of tables; the areas of the table that differ in function will act as classes: heading, cell with a name, cell with data, empty cell.

• The KSFM Journal of Fluid Machinery
• /
• v.12 no.3
• /
• pp.7-12
• /
• 2009

In this paper, shape optimization based on three-dimensional flow analysis has been performed for impeller design of centrifugal compressor. To evaluate the objective function of an isentropic efficiency, Reynolds-averaged Navier-Stokes equations are solved with SST (Shear Stress Transport) turbulence model. The governing equations are discretized by finite volume approximations. The optimization techniques based on the radial basis neural network method are used for the optimization. Latin hypercube sampling as design of experiments is used to generate thirty design points within design space. Sequential quadratic programming is used to search the optimal point based on the radial basis neural network model. Four geometrical variables concerning impeller shape are selected as design variables. The results show that the isentropic efficiency is enhanced effectively from the shape optimization by the radial basis neural network method.

• IE interfaces
• /
• v.25 no.3
• /
• pp.309-318
• /
• 2012

Forecasting probability or likelihood of technology development success has been a crucial factor for critical decisions in technology management such as R&D project selection and go or no-go decision of new product development (NPD) projects. This paper proposes an analytic network process (ANP) approach to forecasting of technology development success. Reviewing literature on factors affecting technology development success has constructed the ANP model composed of four criteria clusters : R&D characteristics, R&D competency, technological characteristics, and technological environment. An alternative cluster comprised of two elements, success and failure is also included in the model. The working of the proposed approach is provided with the help of a case study example of MRAM (magnetic random access memory) technology.

• Proceedings of the KIEE Conference
• /
• 1992.07a
• /
• pp.459-461
• /
• 1992

This paper presents the new method to recognize the 2D patterns dynamically by rotating the input patterns according to the difference vector. Generally neural network with many patterns leads to various recognition ratio. The dynamic management of input patterns means that we can move pixels to desired locations controlled by the difference vector. We divide dual neural network model into two parts at learning phase, respectively. And then we combine them to construct the total network. Our model has some good results such that it has less number of patterns and reduced learning time. At present, we only discuss the four way movement of input patterns. The research for the complex movement will be fulfilled later.

• Structural Monitoring and Maintenance
• /
• v.5 no.1
• /
• pp.51-65
• /
• 2018

A number of sensing techniques have been implemented for detecting defects in civil infrastructures instead of onsite human inspections in structural health monitoring. However, the issue of faults in sensors has not received much attention. This issue may lead to incorrect interpretation of data and false alarms. To overcome these challenges, this article presents a deep learning-based method with a new architecture of Stateful Long Short Term Memory Neural Networks (S-LSTM NN) for detecting sensor fault without going into details of the fault features. As LSTMs are capable of learning data features automatically, and the proposed method works without an accurate mathematical model. The detection of four types of sensor faults are studied in this paper. Non-stationary acceleration responses of a three-span continuous bridge when under operational conditions are studied. A deep network model is applied to the measured bridge data with estimation to detect the sensor fault. Another set of sensor output data is used to supervise the network parameters and backpropagation algorithm to fine tune the parameters to establish a deep self-coding network model. The response residuals between the true value and the predicted value of the deep S-LSTM network was statistically analyzed to determine the fault threshold of sensor. Experimental study with a cable-stayed bridge further indicated that the proposed method is robust in the detection of the sensor fault.

• Nuclear Engineering and Technology
• /
• v.52 no.11
• /
• pp.2565-2571
• /
• 2020

Background: Understanding the behaviour of nuclear fuel claddings by conducting burst test on single cladding tube under simulated loss-of-coolant accident conditions and developing theoretical cum empirical predictive computer codes have been the focus of several investigations. The developed burst criterion (a) assumes symmetrical deformation of cladding tube in contrast to experimental observation (b) interpolates the properties of Zircaloy-4 cladding in mixed α+β phase (c) does not account for azimuthal temperature variations. In order to overcome all these drawbacks of burst criterion, it is reasoned that artificial intelligence technique may be a better option to predict the burst parameters. Methods: Artificial neural network models based on feedforward backpropagation algorithm with logsig transfer function are developed. Results: Neural network architecture of 2-4-4-3, that is model with two hidden layers having four nodes in each layer is found to be the most suitable. The mean, maximum, and minimum prediction errors for this optimised model are 0.82%, 19.62%, and 0.004%, respectively. Conclusion: The burst stress, burst temperature, and burst strain obtained from burst criterion have average deviation of 19%, 12%, and 53% respectively whereas the developed neural network model predicted these parameters with average deviation of 6%, 2%, and 8%, respectively.