• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.20 no.7
• /
• pp.72-79
• /
• 2021

This study reports on the experiment of using deep learning algorithms to determine the machining process of aluminium and steel. A face cutting milling tool was used for machining and the cutting speed was set between 3 and 4 mm/s. Both materials were machined with a depth to 0.5mm and 1.0mm. To demonstrate the developed deep learning algorithm, simulation experiments were performed using the VGGish algorithm in MATLAB toobox. Downcutting was used to cut aluminum and steel as a machining process for high quality and precise learning. As a result of learning algorithms using audio data, 61%-99% accuracy was obtained in four categories: Al 0.5mm, Al 1.0mm, Steel 0.5mm and Steel 1.0mm. Audio discrimination using deep learning is derived as a probabilistic result.

• Tribology and Lubricants
• /
• v.38 no.4
• /
• pp.128-135
• /
• 2022

In the product development process, the reliability of the product can be secured through durability tests. However, since the durability test method is expensive and time consuming, a method to save time and money by utilizing virtual product development (VPD) is required. However, research on the accuracy of the results of virtual product development is required. In this paper, an accelerated durability test was designed and conducted using a planetary gear decelerator. And an analysis model under the same conditions was created and simulated. To correlate the results of the experiment with the results of the analytical model, created a model that can discriminate the wear region using one of the data mining methods, the k-means algorithm method and HSV (Hue, Saturation, Value). The wear area is compared by counting the number of pixels defined as wear through a discrimination model. A similar ratio was calculated by comparing the pixel ratio of the area determined as wear in the entire area. It showed a similar ratio of about 70%, and it is necessary to improve the discrimination method.

• Proceedings of the Korean Society for Agricultural Machinery Conference
• /
• 1996.06c
• /
• pp.823-833
• /
• 1996

An ultimate purpose of this study is to develop an automatic brown rice quality inspection system using image processing technique. In this study emphasis was put on developing an algorithm for discriminating the brown rice kernels depending on their external quality with a color image processing system equipped with an adaptor for magnifying the input image and optical fiber for oblique illumination. Primarily , geometrical and optical features of sample images were analyzed with unhulled paddy and various brown rice kernel samples such as sound, cracked, green-transparent , green-opaque, colored, white-opaque and brokens. Secondary, an algorithm for discrimination of the rice kernels in static state was developed on the basis of the geometrical and optical parameters screened by a statistical analysis(STEPWISE and DISCRIM Procedure, SAS ver.6). Brown rice samples could be discriminated by the algorithm developed in this study with an accuracy of 90% to 96% for the sound , cracked, colored, broken and unhulled , about 81% for the green-transparent and the white-opaque and about 75% for the green-opaque, respectively. A total computing time required for classification was about 100 seconds/1000 kernels with the PC 80486-DX2, 66MHz.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.11
• /
• pp.4426-4442
• /
• 2020

Since scene segmentation is becoming a hot topic in the field of autonomous driving and medical image analysis, researchers are actively trying new methods to improve segmentation accuracy. At present, the main issues in image semantic segmentation are intra-class inconsistency and inter-class indistinction. From our analysis, the lack of global information as well as macroscopic discrimination on the object are the two main reasons. In this paper, we propose a Densely connected residual Attention network (DA-Res2Net) which consists of a dense residual network and channel attention guidance module to deal with these problems and improve the accuracy of image segmentation. Specifically, in order to make the extracted features equipped with stronger multi-scale characteristics, a densely connected residual network is proposed as a feature extractor. Furthermore, to improve the representativeness of each channel feature, we design a Channel-Attention-Guide module to make the model focusing on the high-level semantic features and low-level location features simultaneously. Experimental results show that the method achieves significant performance on various datasets. Compared to other state-of-the-art methods, the proposed method reaches the mean IOU accuracy of 83.2% on PASCAL VOC 2012 and 79.7% on Cityscapes dataset, respectively.

• Journal of Information Processing Systems
• /
• v.19 no.4
• /
• pp.527-539
• /
• 2023

In the positioning algorithm of two-dimensional planar sensor array, the estimation error of time difference-ofarrival (TDOA) algorithm is difficult to avoid. Thus, how to achieve accurate positioning is a key problem of the positioning technology based on planar array. In this paper, a method of sensor reliability discrimination is proposed, which is the foundation for selecting positioning sensors with small error and excellent performance, simplifying algorithm, and improving positioning accuracy. Then, a positioning model is established. The estimation characteristics of the least square method are fully utilized to calculate and fuse the positioning results, and the non-uniform weighting method is used to correct the weighting factors. It effectively handles the decreased positioning accuracy due to measurement errors, and ensures that the algorithm performance is improved significantly. Finally, the characteristics of the improved algorithm are compared with those of other algorithms. The experiment data demonstrate that the algorithm is better than the standard least square method and can improve the positioning accuracy effectively, which is suitable for vibration detection with large noise interference.

• Korean Journal of Remote Sensing
• /
• v.28 no.5
• /
• pp.489-499
• /
• 2012

In this paper, a classifier ensemble framework for remote sensing data classification is presented that combines classification results generated from both different training sets and different classifiers. A core part of the presented framework is to increase a diversity between classification results by using both different training sets and classifiers to improve classification accuracy. First, different training sets that have different sampling densities are generated and used as inputs for supervised classification using different classifiers that show different discrimination capabilities. Then several preliminary classification results are combined via a majority voting scheme to generate a final classification result. A case study of land-cover classification using multi-temporal ENVISAT ASAR data sets is carried out to illustrate the potential of the presented classification framework. In the case study, nine classification results were combined that were generated by using three different training sets and three different classifiers including maximum likelihood classifier, multi-layer perceptron classifier, and support vector machine. The case study results showed that complementary information on the discrimination of land-cover classes of interest would be extracted within the proposed framework and the best classification accuracy was obtained. When comparing different combinations, to combine any classification results where the diversity of the classifiers is not great didn't show an improvement of classification accuracy. Thus, it is recommended to ensure the greater diversity between classifiers in the design of multiple classifier systems.

• Convergence Security Journal
• /
• v.23 no.1
• /
• pp.27-33
• /
• 2023

Digitalization due to COVID-19 has rapidly developed artificial intelligence-based voice recognition technology. However, this technology causes unfair social problems, such as race and gender discrimination if datasets are biased against some groups, and degrades the reliability and security of artificial intelligence services. In this work, we compare and analyze accuracy-based unfairness in biased data environments using VGGNet (Visual Geometry Group Network), ResNet (Residual Neural Network), and MobileNet, which are representative CNN (Convolutional Neural Network) models of artificial intelligence. Experimental results show that ResNet34 showed the highest accuracy for women and men at 91% and 89.9%in Top1-accuracy, while ResNet18 showed the slightest accuracy difference between genders at 1.8%. The difference in accuracy between genders by model causes differences in service quality and unfair results between men and women when using the service.

• Korean Journal of Food Science and Technology
• /
• v.49 no.4
• /
• pp.349-354
• /
• 2017

The geographical origins of raw ginseng (RG) were discriminated using an electronic tongue. Taste screening, DFA (discriminant function analysis), and CDA (canonical discriminant analysis) were used to statistically analyze the data. The taste profile patterns of umami, bitterness, and sweetness of the Korean RG was different from those of the Chinese RG. The Korean RG was stronger than the Chinese RG regarding the taste of umami. DFA discriminated the geographical origins of 154 samples, with a few overlapping samples, between the Korean and Chinese RG. CDA showed that the accuracy of origin discrimination for the Korean and Chinese RGs were 87.01 and 94.81%, respectively. The final accuracy of origin discrimination was 90.91%. The distance between the centroids of each group was 2.7463. Thus, the electronic tongue analysis can be used to efficiently differentiate the geographical origins of RG.

• Journal of the Korean Data and Information Science Society
• /
• v.22 no.4
• /
• pp.619-630
• /
• 2011

In order to estimate an appropriate threshold and evaluate its performance for the data mixed with two different distributions, nine kinds of well-known classification accuracy measures such as MVD, Youden's index, the closest-to- (0,1) criterion, the amended closest-to- (0,1) criterion, SSS, symmetry point, accuracy area, TA, TR are clustered into five categories on the basis of their characters. In credit evaluation study, it is assumed that the score random variable follows normal mixture distributions of the default and non-default states. For various normal mixtures, optimal cut-off points for classification measures belong to each category are obtained and type I and II error rates corresponding to these cut-off points are calculated. Then we explore the cases when these error rates are minimized. If normal mixtures might be estimated for these kinds of real data, we could make use of results of this study to select the best classification accuracy measure which has the minimum error rate.

• Korean Journal of Radiology
• /
• v.22 no.3
• /
• pp.442-453
• /
• 2021

Artificial intelligence (AI) will likely affect various fields of medicine. This article aims to explain the fundamental principles of clinical validation, device approval, and insurance coverage decisions of AI algorithms for medical diagnosis and prediction. Discrimination accuracy of AI algorithms is often evaluated with the Dice similarity coefficient, sensitivity, specificity, and traditional or free-response receiver operating characteristic curves. Calibration accuracy should also be assessed, especially for algorithms that provide probabilities to users. As current AI algorithms have limited generalizability to real-world practice, clinical validation of AI should put it to proper external testing and assisting roles. External testing could adopt diagnostic case-control or diagnostic cohort designs. A diagnostic case-control study evaluates the technical validity/accuracy of AI while the latter tests the clinical validity/accuracy of AI in samples representing target patients in real-world clinical scenarios. Ultimate clinical validation of AI requires evaluations of its impact on patient outcomes, referred to as clinical utility, and for which randomized clinical trials are ideal. Device approval of AI is typically granted with proof of technical validity/accuracy and thus does not intend to directly indicate if AI is beneficial for patient care or if it improves patient outcomes. Neither can it categorically address the issue of limited generalizability of AI. After achieving device approval, it is up to medical professionals to determine if the approved AI algorithms are beneficial for real-world patient care. Insurance coverage decisions generally require a demonstration of clinical utility that the use of AI has improved patient outcomes.