• Journal of the Korean Society of Radiology
• /
• v.10 no.8
• /
• pp.637-643
• /
• 2016

Due to the popularity of artificial intelligent, medical image processing using artificial neural network is increasingly attracting the attention of academic and industry researches. Deep learning with a convolutional neural network has been proved to very effective representation of images. However, the training process requires high performance H/W platform. Thus, the realtime learning of a large number of high dimensional samples within low-power devices is a challenging problem. In this paper, we attempt to establish this possibility by presenting a realtime neural network method on Raspberry pi using online sequential extreme learning machine. Our experiments on high-dimensional dataset show that the proposed method records an almost real-time execution.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.37 no.11
• /
• pp.1397-1405
• /
• 2013

The optimal rib pattern design of B-pillar trim considering occupant head protection can be determined by two methods. One is the conventional approximate optimization method that uses the statistical design of experiments (DOE) and response surface method (RSM). Generally, approximated optimum results are obtained through the iterative process by trial-and-error. The quality of results strongly depends on the factors and levels assigned by a designer. The other is a methodology derived from previous work by the authors, called the sequential design of experiments (SDOE), to reduce the trial-and-error procedure and to find an appropriate condition for using artificial neural network (ANN) systematically. An appropriate condition is determined from the iterative process based on the analysis of means. With this new technique and ANN, it is possible to find an optimum design accurately and efficiently.

• Journal of Intelligence and Information Systems
• /
• v.23 no.1
• /
• pp.109-126
• /
• 2017

In the various kinds of intelligent services, many studies for detecting emotion are in progress. Particularly, studies on emotion recognition at the particular time have been conducted in order to provide personalized experiences to the audience in the field of exhibition though facial expressions change as time passes. So, the aim of this paper is to build a model to predict the audience's emotion from the changes of facial expressions while watching an exhibit. The proposed model is based on both sequential neural network and the Valence-Arousal model. To validate the usefulness of the proposed model, we performed an experiment to compare the proposed model with the standard neural-network-based model to compare their performance. The results confirmed that the proposed model considering time sequence had better prediction accuracy.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.7 no.8
• /
• pp.1802-1808
• /
• 2003

In this paper, we propose a method about the extraction of the pedestrian tracking trajectory in the road and we used the method of mosaic background extraction and adaptive neural network for automatic pedestrian tracking system. We used mosaic background extraction to overcome ghost phenomenon. And we detected pedestrian using differential image analysis. We used adaptive neural network for multiple pedestrian tracking that non­rigid form moving. The ART2 network is capable of detecting the mass­centers of moving objects within one frame. The history of neurons positions in the sequential frames approximates the traces of the targets. The experiments done with the network in simulated environment show promising results.

• 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.

• 제어로봇시스템학회:학술대회논문집
• /
• 2005.06a
• /
• pp.287-289
• /
• 2005

This paper presents a hybrid neural network system which used a Self-Organizing Map and Multilayer Perceptron for the problem of Malay syllables speech recognition. The novel idea in this system is the usage of a two-dimension Self-organizing feature map as a sequential mapping function which transform the phonetic similarities or acoustic vector sequences of the speech frame into trajectories in a square matrix where elements take on binary values. This property simplifies the classification task. An MLP is then used to classify the trajectories that each syllable in the vocabulary corresponds to. The system performance was evaluated for recognition of 15 Malay common syllables. The overall performance of the recognizer showed to be 91.8%.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.32B no.12
• /
• pp.1581-1590
• /
• 1995

Recently, advanced video coding techniques using segmentation technique have been actively researched as candidates for video coding of MPEG-4 standard. The conventional segmentation techniques are unsuitable for real-time process because they have sequential structure. In this paper, we propose a new image segmentation technique using competitive learning neural network for vector quantization. The proposed segmentation procedure consist of prefiltering, primary and secondary segmentation, and a small region ellimination process. Primary segmentation segments input image in detail. Secondary segmentation merges similar region using a repetitive FSCL(Frequency sensitive competive learning) neural network. In this process, it is possible to segment an image from high resolution to low resolution by adjusting the number of repetition. Finally, small regions are merged into adjacent regions. Experimental results show that the procedure described yields reconstructed images of reasonably acceptable quality at bit rates of 0. 25 - 0.3 bit/pel.

• Journal of Institute of Control, Robotics and Systems
• /
• v.4 no.3
• /
• pp.378-383
• /
• 1998

A FOV, that stands for "Field Of View", refers to the maximum area where a camera could be wholly seen. If a FOV of CCD camera cannot the cover overall inspection area, the overall inspection area should be divided into sub-areas of size FOV. In this paper, we propose a new neural network-based FOV generation method by using a newly modified self-organizing map(SOM) which has multiple structure based on a self-organizing map, and uses new training rule that is composed of the movement, creation and deletion terms. Then, experiment results using real PCB indicate the superiority of the method developed in this study to the existing sequential method.al method.

• ETRI Journal
• /
• v.41 no.5
• /
• pp.560-573
• /
• 2019

Two supervised learning algorithms, a basic neural network and a long short-term memory recurrent neural network, are applied to traffic including DDoS attacks. The joint effects of preprocessing methods and hyperparameters for machine learning on performance are investigated. Values representing attack characteristics are extracted from datasets and preprocessed by two methods. Binary classification and two optimizers are used. Some hyperparameters are obtained exhaustively for fast and accurate detection, while others are fixed with constants to account for performance and data characteristics. An experiment is performed via TensorFlow on three traffic datasets. Three scenarios are considered to investigate the effects of learning former traffic on sequential traffic analysis and the effects of learning one dataset on application to another dataset, and determine whether the algorithms can be used for recent attack traffic. Experimental results show that the used preprocessing methods, neural network architectures and hyperparameters, and the optimizers are appropriate for DDoS attack detection. The obtained results provide a criterion for the detection accuracy of attacks.

• Journal of the Semiconductor & Display Technology
• /
• v.22 no.2
• /
• pp.128-132
• /
• 2023

Recently, with the development of deep learning technology, research on pedestrian attribute recognition technology using deep neural networks has been actively conducted. Existing pedestrian attribute recognition techniques can be obtained in such a way as global-based, regional-area-based, visual attention-based, sequential prediction-based, and newly designed loss function-based, depending on how pedestrian attributes are detected. It is known that the performance of these pedestrian attribute recognition technologies varies greatly depending on the type of backbone network that constitutes the deep neural networks model. Therefore, in this paper, several backbone networks are applied to the baseline pedestrian attribute recognition model and the performance changes of the model are analyzed. In this paper, the analysis is conducted using Resnet34, Resnet50, Resnet101, Swin-tiny, and Swinv2-tiny, which are representative backbone networks used in the fields of image classification, object detection, etc. Furthermore, this paper analyzes the change in time complexity when inferencing each backbone network using a CPU and a GPU.