• Applied Biological Chemistry
• /
• v.20 no.3
• /
• pp.279-284
• /
• 1977

Like that mentioned in the 2nd report, because of panel's sense of psychological and physiological responsibility caused by multi-samples, great errors in experimental results are expected. So as to cut down these errors, the new method called "Pair Comparision with Standard" that reduces test frequency and is superior in detecting power is designed, and its mathematical model is proposed. This paper suggests that this method can be used for screening test that, first of all, selects 4-5 of multi-samples and the most efficient sensory evaluation method in laboratorial quality study is that, after screening by this method, Trio Paired Comparison for the final justification is applicated.

• Journal of Broadcast Engineering
• /
• v.21 no.4
• /
• pp.493-505
• /
• 2016

Among modern computer vision and video surveillance systems, the pedestrian counting system is a one of important systems in terms of security, scheduling and advertising. In the field of, pedestrian counting remains a variety of challenges such as changes in illumination, partial occlusion, overlap and people detection. During pedestrian counting process, the biggest problem is occlusion effect in crowded environment. Occlusion and overlap must be resolved for accurate people counting. In this paper, we propose a novel pedestrian counting system which improves existing pedestrian tracking method. Unlike existing pedestrian tracking method, proposed method shows that average filter tracking method can improve tracking performance. Also proposed method improves tracking performance through frame compensation and outlier removal. At the same time, we keep various information of tracking objects. The proposed method improves counting accuracy and reduces error rate about S6 dataset and S7 dataset. Also our system provides real time detection at the rate of 80 fps.

• Journal of the Mineralogical Society of Korea
• /
• v.19 no.3 s.49
• /
• pp.189-195
• /
• 2006

Compression work on a pyrite powder has been carried out using energy dispersive X-ray diffraction (EDXRD) with Mao-Bell type diamond anvil cell (DAC) and synchrotron radiation(SR) at room temperature. It has been reported the bulk moduli of pyrite show the large variations depending on the experimental conditions as well as the apparatus used. Thus, two kinds of sample in different pressure transmitting media of both NaCl and MgO powder emerged in alcoholic fluids were subjected to measure their compressibilities. Bulk moduli thus obtained are 138.9 GPa and 198.2 GPa, respectively, and this result contradicts to the anticipated values according to the hydrostaticity conditions of the sample chamber. This might be due to the alcoholic fluids phase transition mainly with the side effects from the difference of both solid state detector (SSD) used and E*d value applied. All experiments were performed at the Beam Line 1B2 of Pohang Light Source (PLS).

• The Journal of the Acoustical Society of Korea
• /
• v.39 no.5
• /
• pp.414-423
• /
• 2020

In this paper, we propose a Dilated Convolution Gate Linear Unit (DCGLU) to mitigate the lack of sparsity and small receptive field problems caused by the segmentation map extraction process in sound event detection with weak labels. In the advent of deep learning framework, segmentation map extraction approaches have shown improved performance in noisy environments. However, these methods are forced to maintain the size of the feature map to extract the segmentation map as the model would be constructed without a pooling operation. As a result, the performance of these methods is deteriorated with a lack of sparsity and a small receptive field. To mitigate these problems, we utilize GLU to control the flow of information and Dilated Convolutional Neural Networks (DCNNs) to increase the receptive field without additional learning parameters. For the performance evaluation, we employ a URBAN-SED and self-organized bird sound dataset. The relevant experiments show that our proposed DCGLU model outperforms over other baselines. In particular, our method is shown to exhibit robustness against nature sound noises with three Signal to Noise Ratio (SNR) levels (20 dB, 10 dB and 0 dB).

• Journal of KIISE
• /
• v.43 no.12
• /
• pp.1356-1364
• /
• 2016

This paper proposes a novel aesthetic photo recomposition method using a deep convolutional neural network (DCNN). Previous recomposition approaches define the aesthetic score of photo composition based on the distribution of salient objects, and enhance the photo composition by maximizing the score. These methods suffer from heavy computational overheads, and often fail to enhance the composition because their optimization depends on the performance of existing salient object detection algorithms. Unlike previous approaches, we address the photo recomposition problem by utilizing DCNN, which shows remarkable performance in object detection and recognition. DCNN is used to iteratively predict cropping directions for a given photo, thus generating an aesthetically enhanced photo in terms of composition. Experimental results and user study show that the proposed framework can automatically crop the photo to follow specific composition guidelines, such as the rule of thirds.

• Korean Journal of Plant Resources
• /
• v.20 no.5
• /
• pp.437-441
• /
• 2007

The argE gene of E.coli was introduced into #Ilmibyeo# cultivar of rice by Agrobacterium tumefaciens and a large number of transgenic plants were produced. Embryogenic calli were co-cultivated with A. tumefaciens strain AGL1 carrying the plasmid pCAMBIA1300 containing hygromycin resistance(HygR). Transgenic plants showing in vitro resistance to 50mg/L hygromycin were obtained using a selection procedure. Stable integration of argE and HPT genes into chromosomal DNA was proven by southern blot analysis and PCR analysis of genomic isolated from $T_0$ progenies. The fragments of 650 bp(HPT) were detected in transgenic rice lines. The 230 bp(argE) fragments were showed in agarose gel, and detected fragments were matched with size of argE specific primer. The microscopic feature of leaf on scanning electron microscope(SEM) revealed differences between clear and chalky in shape and arrangement of stoma but did not discriminate.

• Asian Journal of Turfgrass Science
• /
• v.23 no.2
• /
• pp.331-344
• /
• 2009

Buffalograss is an important turfgrass species with excellent cold, heat, and drought tolerance. Understanding the physiological integration of buffalograss under heterogeneous conditions helps to develop cultural practices that better use limited resources for uniform turf quality. The objective of this study was to evaluate physiological integration of buffalograss under water deficit stress and the involvement of lipid peroxidation and antioxidants in the process. In one experiment, buffalograss was planted in the center of a four-compartment growth unit. Watering frequencies, once a week(+) and once in two weeks(-), were combined with the sand(S) or peat(P) in each unit to generate five total treatments(P+S-P-S+, P+P+P+P+, S-S-S-S-, P-P-P-P-, and S+S+S+S+). The average number of shoot established from the heterogeneous root-zone medium was higher than the average of four possible homogeneous media. In second experiment, single ramet in Hoagland solution($S_0$) or single ramet in Hoagland solution with 20% PEG-6000($S_s$) were compared with two connectedramets under different treatments. Treatments for connected ramets were young ramet in Hoagland solution($Y_{os}$) and old ramet in Hoagland solution with 20% PEG-6000($O_{os}$), and old ramet in Hoagland solution($O_{ys}$) and young ramet in Hoagland solution with 20% PEG-6000($Y_{ys}$). Lipid peroxidation, antioxidants, and proline showedphysiological integration between ramets subjected to different levels of water stress. Superoxide dismutase(SOD), Guaiacol peroxidase(G-POD), malondialdehyde(MDA), and free proline also showed different time courses and relative activities during the physiological integration.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.4
• /
• pp.93-103
• /
• 2021

Despite the efforts of financial authorities in conducting the direct management and supervision of collection agents and bond-collecting guideline, the illegal and unfair collection of debts still exist. To effectively prevent such illegal and unfair debt collection activities, we need a method for strengthening the monitoring of illegal collection activities even with little manpower using technologies such as unstructured data machine learning. In this study, we propose a classification model for illegal debt collection that combine machine learning such as Support Vector Machine (SVM) with a rule-based technique that obtains the collection transcript of loan companies and converts them into text data to identify illegal activities. Moreover, the study also compares how accurate identification was made in accordance with the machine learning algorithm. The study shows that a case of using the combination of the rule-based illegal rules and machine learning for classification has higher accuracy than the classification model of the previous study that applied only machine learning. This study is the first attempt to classify illegalities by combining rule-based illegal detection rules with machine learning. If further research will be conducted to improve the model's completeness, it will greatly contribute in preventing consumer damage from illegal debt collection activities.

• Journal of Food Hygiene and Safety
• /
• v.38 no.5
• /
• pp.297-304
• /
• 2023

Domoic acid (DA), a neurotoxin produced naturally by diatoms, is responsible for incidents of amnesic shellfish poisoning. In this study, a modified analytical method was established to determine domoic acid in seafood using solid phase extraction cleanup and optimizing the amount of sample and extraction solvent to reduce interference effects. The modified method using high-performance liquid chromatography with ultraviolet detection was validated using three seafood matrices (mussel, red snow crab, and anchovy) at three concentrations (1, 2, and 4 mg/kg) and compared to the Food Code method. Compared to the Food Code method, the modified method showed better performance in terms of linearity (R²>0.999), detection limit (0.02-0.03 mg/kg), quantification limit (0.05-0.09 mg/kg), intra-/inter-day accuracy (86.2-100.4%), and intra-/inter-day precision (0.2-4.0%). Furthermore, the method was successfully applied for the analysis of 87 seafood samples marketed in Korea, and DA was detected at a low concentration of 140 ㎍/kg in one anchovy sample. These results suggest that the modified method can be used for routine determination of DA in seafood.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.52 no.9
• /
• pp.13-18
• /
• 2015

The mutilevel holographic data storage offers considerable advantage for capacity, since it can store more than one bit per pixel. In this paper, we search the number of codewords for each code depending on three conditions: (1) the number of levels, (2) the number of pixels in a codeword, and (3) the minimum Euclidean distance of a code. Increasing the number of levels per pixel creates more capacity, while causing more errors, by reducing the noise margin. Increasing the number of pixels in a codeword can increase the code rate, which means more capacity, but increases the complexity of the encoder/decoder of the code. Increasing the minimum distance of a code reduces the detection error, while reducing the code rate of the code. In such a fashion, a system design will always have pros and cons, but our task is to find out an effective one under the given conditions for the system requirements. Therefore, the numbers we searched can provide some guidelines for effective code design.