• Journal of Nutrition and Health
• /
• v.47 no.3
• /
• pp.193-205
• /
• 2014

Purpose: This study was conducted in order to investigate nutrient and food intake status and dietary quality in Korean adults according to nutrition label utilization. Methods: We analyzed data from the combined 2010-2011 KNHANES (Korean National Health and Nutrition Examination Survey). The analysis included 8190 adults aged 19 to 64 years. In this study, according to nutrition label utilization, we classified the subjects according to the "non-utilization of nutrition label (NUNL)" group (male, n = 2716, female, n = 3147), "identification of nutrition label (INL)" group (male, n = 143, female, n = 330), and "Utilization of nutrition label (UNL)" group (male, n = 363, female, n = 1491). Nutrient and food group intake, NAR (nutrient adequacy ratio), MAR (mean adequacy ratio), and INQ (index of nutritional quality) were analyzed using data from the 24-recall method. Results: Results of this study showed that subjects in the NUNL group were significantly more likely to drink alcohol compared with the other two groups. The NUNL group showed a significantly higher frequency of consuming instant noodles, Soju (male), and carbonated drink (female) than the UNL group, whereas the NUNL group showed a significantly lower frequency of consuming milk, soymilk, and yogurt than the UNL group. In addition, regarding diet quality (NAR and INQ), significantly lower vitamin $B_2$, vitamin C, and calcium was observed in the NUNL group compared with the UNL group. For both male and female, significantly higher MAR was observed in the UNL group than in the NUNL group. The NUNL group showed significantly lower consumption of milk compared to the UNL group. Conclusion: Good dietary practice such as referring to nutrition labels and its influence can affect the quality of nutritional intake and selection of food, while it can also provide basic data for specific nutrition education regarding use of nutrition labeling.

• Journal of Food Hygiene and Safety
• /
• v.34 no.5
• /
• pp.502-507
• /
• 2019

Cephalopods are one of the most important fishery resources in the world because of their desirable taste and nutritional value. In south Korea, one of the countries in which a large amount of seafood is consumed, cephalopods (e.g., octopus, squid, and cuttlefish) have an annual consumption rate of over 400,000 metric tons. In this study, octopus and squid products (n=28) sold on the market were monitored by analyzing sequences of DNA barcode markers (cytochrome c oxidase subunit I and 16S ribosomal RNA genes). For species identification, the NCBI BLAST database was screened with the sequences and analyzed as a query. In this BLAST search, twelve squid products showed 99-100% sequence identity to Dosidicus gigas (n=3) and Todarodes pacificus (n=9). In the case of the other 16 products that were declared using octopus as raw materials on the labels, six products were identified as Cistopus taiwanicus (n=1), Amphioctopus marginatus (n=1), Scaeurgus unicirrhus (n=1), and Dosidicus gigas (n=3). Monitoring results indicated that a significant percentage (37.5%) of mislabeling was present in octopus products sold on the South Korean market.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.24 no.6
• /
• pp.583-598
• /
• 2022

As human-based tunnel inspections are affected by the subjective judgment of the inspector, making continuous history management difficult. There is a lot of deep learning-based automatic crack detection research recently. However, the large public crack datasets used in most studies differ significantly from those in tunnels. Also, additional work is required to build sophisticated crack labels in current tunnel evaluation. Therefore, we present a method to improve crack detection performance by inputting existing datasets into a deep learning model. We evaluate and compare the performance of deep learning models trained by combining existing tunnel datasets, high-quality tunnel datasets, and public crack datasets. As a result, DeepLabv3+ with Cross-Entropy loss function performed best when trained on both public datasets, patchwise classification, and oversampled tunnel datasets. In the future, we expect to contribute to establishing a plan to efficiently utilize the tunnel image acquisition system's data for deep learning model learning.

• The Journal of the Acoustical Society of Korea
• /
• v.40 no.5
• /
• pp.479-487
• /
• 2021

In this paper, we propose an Sound Event Detection (SED) model using self-training based on a noisy student model. The proposed SED model consists of two stages. In the first stage, a mean-teacher model based on an Residual Convolutional Recurrent Neural Network (RCRNN) is constructed to provide target labels regarding weakly labeled or unlabeled data. In the second stage, a self-training-based noisy student model is constructed by applying different noise types. That is, feature noises, such as time-frequency shift, mixup, SpecAugment, and dropout-based model noise are used here. In addition, a semi-supervised loss function is applied to train the noisy student model, which acts as label noise injection. The performance of the proposed SED model is evaluated on the validation set of the Detection and Classification of Acoustic Scenes and Events (DCASE) 2020 Challenge Task 4. The experiments show that the single model and ensemble model of the proposed SED based on the noisy student model improve F1-score by 4.6 % and 3.4 % compared to the top-ranked model in DCASE 2020 challenge Task 4, respectively.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.20 no.2
• /
• pp.95-111
• /
• 2021

As the population decreases in an aging society, the average age of drivers increases. Accordingly, the elderly at high risk of being in an accident need autonomous-driving vehicles. In order to secure driving safety on the road, several technologies to respond to various obstacles are required in those vehicles. Among them, technology is required to recognize static obstacles, such as poor road conditions, as well as dynamic obstacles, such as vehicles, bicycles, and people, that may be encountered while driving. In this study, we propose a deep neural network algorithm capable of simultaneously detecting these two types of obstacle. For this algorithm, we used 1,418 road images and produced annotation data that marks seven categories of dynamic obstacles and labels images to indicate road damage. As a result of training, dynamic obstacles were detected with an average accuracy of 46.22%, and road surface damage was detected with a mean intersection over union of 74.71%. In addition, the average elapsed time required to process a single image is 89ms, and this algorithm is suitable for personal mobility vehicles that are slower than ordinary vehicles. In the future, it is expected that driving safety with personal mobility vehicles will be improved by utilizing technology that detects road obstacles.

• The Journal of the Korea Contents Association
• /
• v.19 no.3
• /
• pp.510-519
• /
• 2019

The purpose of this study was to analyze the nutrition labeling of Guk, Tang, Jjigae HMR products to provide consumers with appropriate information when selecting products, and to provide basic data on the national policies. In this study, nutritional labels of 176 products were analyzed with 57 Guk, 75 Tang, 44 Jjigae. In the menu frequency of products, Guk has the products of the specific purposes, Tang has animal protein product, Jjigae has popular products. As a result of comparing the portion size and 9 major nutrients of the product, the average sodium content of Guk was 1558.5 mg, Tang was 1472.3mg, Jjigae was 2118.0mg. By the storage temperature, the average sodium content of HMR product was 2022.9mg in freezing(below $-18^{\circ}C$), 1676.7mg in cold($-2{\sim}10^{\circ}C$), and 1250.9mg in room temperature($1{\sim}35^{\circ}C$). Therefore, it is necessary to focus on the sodium content of Frozen products in the attempt of reducing sodium in HMR products.

• Korean Journal of Community Nutrition
• /
• v.24 no.2
• /
• pp.137-151
• /
• 2019

Objectives: The rapidly changing dietary environment requires a study that addresses the status of middle and high school students regarding their consumption of convenience food sold at convenience stores. Methods: This study examined adolescents' lifestyle patterns, dietary habits, and status of consuming convenience food at convenience stores. A total of 659 students (329 middle school students and 330 high school students) in Incheon participated in this questionnaire survey. Results: The mean age of the subjects was 13.7 years for the middle school students, and 16.6 years for the high school students. The gender and grade distributions in the middle and high school students were similar. The middle school students reported that they spent more time using electronic devices (p<0.001) or watching TV (p<0.001) than high school students. More than 60% of middle and high school students consumed convenience food at convenience stores without statistical difference between the two groups. The main reason for consuming convenience food from convenience stores was its convenience followed by taste in both groups. Despite the high frequency of consuming convenience food, the students rarely checked the nutrition labels at the time of purchase. On the other hand, they were still most concerned about the nutritional value of the convenience foods when they consumed convenience foods. The most frequently consumed convenience food was ramyon in both groups. Significant positive correlations were observed between the frequency of consuming convenience food at convenience stores and lifestyle factors for the middle school students, including monthly allowance, time for using electronic devices, and number of private lessons. For the high school students, however, the only monthly allowance had a significant positive correlation with the consumption. Conclusions: Adolescents are increasingly exposed to convenience foods and relevant nutritional issues are a concern. Therefore, a dietary environment that is adequately formed for the healthy development of youth as well as systematic nutrient education that is appropriately designed for both middle and high school students is required.

• Journal of Korean Academy of Oral Health
• /
• v.42 no.4
• /
• pp.199-203
• /
• 2018

Objectives: The market for bottled water is increasing steadily in South Korea. Bottled water contains several naturally occuring minerals, such as calcium, magnesium, sodium, and fluoride. Fluoride is proven to be effective in preventing dental caries. In South Korea, the maximum permissible concentration of fluoride is 2 ppm for bottled water and 1.5 ppm for tap water. The aim of this study was to investigate the fluoride content of different commercially available brands of bottled water in South Korea, and compare the measured fluoride concentration to the concentration written on the label of each brand of bottled water. Methods: Twenty-seven of the 59 different brands of bottled water produced in South Korea were investigated in this study. Three bottles of each brand were purchased from supermarkets, marts, and convenience stores in each region of Korea in August 2016. For each bottled water brand, the fluoride content was measured three times using a fluoride-ion selective electrode (Orion ionplus Fluoride Electrode 9609, Orion Research, USA). The calibration curve was generated using 0.2 and 2 ppm standard solutions, and confirmed using a 1 ppm standard solution. Results: The mean fluoride content of the 27 brands of bottled water was $0.374{\pm}0.332mg/L$ (range=0.040 to 1.172 mg/L). The fluoride content was labeled by the manufacturer, on each of the tested brands of bottled water. In eight brands, the labeled fluoride content differed from the experimental data. The minimum to maximum fluoride content measured from 10 brands showed a variation of 0.3 mg/L or more when compared to the labeled fluoride content. Conclusions: This study investigated the fluoride content of various brands of bottled water produced in South Korea and compared the measured fluoride levels with fluoride information on the bottle labels. To ensure that consumers are suitably informed regarding their exposure to fluoride, correct labelling of fluoride content in bottled water is important.

• Journal of Broadcast Engineering
• /
• v.24 no.3
• /
• pp.472-484
• /
• 2019

This paper proposes a time-domain sound event detection algorithm using DNN (Deep Neural Network). In this system, time domain sound waveform data which is not converted into the frequency domain is used as input to the DNN. The overall structure uses CRNN structure, and GLU, ResNet, and Squeeze-and-excitation blocks are applied. And proposed structure uses structure that considers features extracted from several layers together. In addition, under the assumption that it is practically difficult to obtain training data with strong labels, this study conducted training using a small number of weakly labeled training data and a large number of unlabeled training data. To efficiently use a small number of training data, the training data applied data augmentation methods such as time stretching, pitch change, DRC (dynamic range compression), and block mixing. Unlabeled data was supplemented with insufficient training data by attaching a pseudo-label. In the case of using the neural network and the data augmentation method proposed in this paper, the sound event detection performance is improved by about 6 %(based on the f-score), compared with the case where the neural network of the CRNN structure is used by training in the conventional method.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.19 no.1
• /
• pp.161-170
• /
• 2019

Multi-product mass production small and medium enterprise factories have a wide variety of products and a large number of products, wasting manpower and expenses for inventory management. In addition, there is no way to check the status of inventory in real time, and it is suffering economic damage due to excess inventory and shortage of stock. There are many ways to build a real-time data collection environment, but most of them are difficult to afford for small and medium-sized companies. Therefore, smart factories of small and medium enterprises are faced with difficult reality and it is hard to find appropriate countermeasures. In this paper, we implemented the contents of extension of existing inventory management method through character extraction on label with barcode and QR code, which are widely adopted as current product management technology, and evaluated the effect. Technically, through preprocessing using OpenCV for automatic recognition and classification of stock labels and barcodes, which is a method for managing input and output of existing products through computer image processing, and OCR (Optical Character Recognition) function of Google vision API. And it is designed to recognize the barcode through Zbar. We propose a method to manage inventory by real-time image recognition through Raspberry Pi without using expensive equipment.