• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 2002년도 추계학술대회 논문집
• /
• pp.1007-1010
• /
• 2002

The accuracy of high speed feeding type laser cutting M/C is the major factor directly concerned with the accuracy of the processed work, and the feed errors of feed system make the machining errors of work directly on processing. In this point, this study focused on the generative elements in feed errors of laser cutting M/C when operating its laser head. In order to improve the accuracy of this machining center, feed errors are measured by a laser interferometer.

• 소성∙가공
• /
• 제23권7호
• /
• pp.446-452
• /
• 2014

In the current study finite element forming analysis is performed to understand the final geometric accuracy limitations for the stamping of an automotive S-rail from four different steel sheets having tensile strengths of 340MPa, 440MPa, 590MPa and 780MPa. Comparisons between the analysis and the experiments for both springback and formability as measured by the amount of edge draw-in and the thickness distribution were conducted. The springback modes were classified according to a scheme proposed in the current investigation and the error was calculated using the normalized root mean square error method. While the analysis results show fairly good agreement with the experimental data for deformation and formability, the simulation accuracy is lower for predicting wall curl, camber and section twist as the UTS of steel sheet increases.

• ETRI Journal
• /
• 제45권4호
• /
• pp.666-677
• /
• 2023

In the last few years, convolutional neural networks (CNNs) have demonstrated good performance while solving various computer vision problems. However, since CNNs exhibit high computational complexity, signal processing is performed on the server side. To reduce the computational complexity of CNNs for edge computing, a lightweight algorithm, such as a MobileNet, is proposed. Although MobileNet is lighter than other CNN models, it commonly achieves lower classification accuracy. Hence, to find a balance between complexity and accuracy, additional hyperparameters for adjusting the size of the model have recently been proposed. However, significantly increasing the number of parameters makes models dense and unsuitable for devices with limited computational resources. In this study, we propose a novel MobileNet architecture, in which the number of parameters is adaptively increased according to the importance of feature maps. We show that our proposed network achieves better classification accuracy with fewer parameters than the conventional MobileNet.

• 한국산업융합학회 논문집
• /
• 제26권6_3호
• /
• pp.1233-1241
• /
• 2023

This paper investigates the efficient application of drone LiDAR technology for acquiring precise point cloud data in construction and civil engineering. A structured workflow encompassing data acquisition, processing, and accuracy verification is introduced. Practical testing on a construction site affirms that drone LiDAR surveying yields accurate and reliable data across various applications. With a focus on accuracy and verification, the results contribute to the progression of surveying methodologies in construction and civil engineering. The findings provide valuable insights into the dynamic technological landscape of these fields, establishing a foundation for more effective and precise surveying techniques. This study underscores the transformative potential of drone LiDAR technology in shaping the future of construction and civil engineering survey practices.

• Journal of Platform Technology
• /
• 제12권1호
• /
• pp.3-11
• /
• 2024

Deaf people who use sign language as their first language sometimes have difficulty communicating because they do not know spoken Korean. Deaf people are also members of society, so we must support to create a society where everyone can live together. In this paper, we present a method to increase the recognition rate of Korean sign language using a CNN model. When the original image was used as input to the CNN model, the accuracy was 0.96, and when the image corresponding to the skin area in the YCbCr color space was used as input, the accuracy was 0.72. It was confirmed that inserting the original image itself would lead to better results. In other studies, the accuracy of the combined Conv1d and LSTM model was 0.92, and the accuracy of the AlexNet model was 0.92. The CNN model proposed in this paper is 0.96 and is proven to be helpful in recognizing Korean sign language.

• Journal of Information Processing Systems
• /
• 제2권1호
• /
• pp.44-47
• /
• 2006

One of the critical issues in a sensor network concerns the detection of changes in data streams. Recently presented change detection schemes primarily use a sliding window model to detect changes. In such a model, a distance function is used to compare two sliding windows. Therefore, the performance of the change detection scheme is greatly influenced by the distance function. With regard to sensor nodes, however, energy consumption constitutes a critical design concern because the change detection scheme is implemented in a sensor node, which is a small battery-powered device. In this paper, we present a comparative study of various distance functions in terms of execution time, energy consumption, and detecting accuracy through simulation of speech signal data. The simulation result demonstrates that the Euclidean distance function has the highest performance while consuming a low amount of power. We believe our work is the first attempt to undertake a comparative study of distance functions in terms of execution time, energy consumption, and accuracy detection.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1996년도 Proceedings of the Korea Automatic Control Conference, 11th (KACC); Pohang, Korea; 24-26 Oct. 1996
• /
• pp.153-156
• /
• 1996

A new filtering algorithm for radar tracking is developed based on the fact that correct evaluation of the measurement error covariance can be made possible by doing it with respect to the Cartesian state vector. The new filter may be viewed as a modification of the extended Kalman filter where the variance of the range measurement errors is evaluated in an adaptive manner. The structure of the proposed filter allows sequential measurement processing scheme to be incorporated into the scheme, and this makes the resulting algorithm favorable in both estimation accuracy and computational efficiency.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2004년도 ICCAS
• /
• pp.1307-1312
• /
• 2004

Word sense disambiguation is one of the most important problems in natural language processing research topics such as information retrieval and machine translation. Many approaches can be employed to resolve word ambiguity with a reasonable degree of accuracy. These strategies are: knowledge-based, corpus-based, and hybrid-based. This paper pays attention to the corpus-based strategy. The purpose of this paper is to compare three famous machine learning techniques, Snow, SVM and Naive Bayes in Word-Sense Disambiguation on Thai language. 10 ambiguous words are selected to test with word and POS features. The results show that SVM algorithm gives the best results in solving of Thai WSD and the accuracy rate is approximately 83-96%.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2003년도 ICCAS
• /
• pp.1727-1731
• /
• 2003

In this paper we proposed backward fuzzy Petri-net to diagnoses faults in power systems by using the time sequence information of protection system. As the complexity of power systems increases, especially in the case of multiple faults or incorrect operation of protective devices, fault diagnosis requires new and systematic methods to the reasoning process, which improves both its accuracy and its efficiency. The fuzzy Petri-net models of protection system are composed of the operating process of protective devices and the fault diagnosis process. Fault diagnosis model, which makes use of the nature of fuzzy Petri-net, is developed to overcome the drawbacks of methods that depend on operator knowledge. The proposed method can reduce processing time and increase accuracy when compared with the traditional methods. And also this method covers online processing of real-time data from SCADA (Supervisory Control and Data Acquisition)

• Journal of Information Processing Systems
• /
• 제12권1호
• /
• pp.109-128
• /
• 2016

This paper presents a new combined forecasting method that is guided by the soft set theory (CFBSS) to predict business failures with different sample sizes. The proposed method combines both qualitative analysis and quantitative analysis to improve forecasting performance. We considered an expert system (ES), logistic regression (LR), and support vector machine (SVM) as forecasting components whose weights are determined by the receiver operating characteristic (ROC) curve. The proposed procedure was applied to real data sets from Chinese listed firms. For performance comparison, single ES, LR, and SVM methods, the combined forecasting method based on equal weights (CFBEWs), the combined forecasting method based on neural networks (CFBNNs), and the combined forecasting method based on rough sets and the D-S theory (CFBRSDS) were also included in the empirical experiment. CFBSS obtains the highest forecasting accuracy and the second-best forecasting stability. The empirical results demonstrate the superior forecasting performance of our method in terms of accuracy and stability.