• International Journal of Internet, Broadcasting and Communication
• /
• v.16 no.1
• /
• pp.124-131
• /
• 2024

Breast cancer classification using ultrasound, while widely employed, faces challenges due to its relatively low predictive value arising from significant overlap in characteristics between benign and malignant lesions, as well as operator-dependency. To alleviate these challenges and reduce dependency on radiologist interpretation, the implementation of automatic breast cancer classification in ultrasound image can be helpful. To deal with this problem, we propose a semi-supervised deep learning framework for breast cancer classification. In the proposed method, we could achieve reasonable performance utilizing less than 50% of the training data for supervised learning in comparison to when we utilized a 100% labeled dataset for training. Though it requires more modification, this methodology may be able to alleviate the time-consuming annotation burden on radiologists by reducing the number of annotation, contributing to a more efficient and effective breast cancer detection process in ultrasound images.

• Journal of KIISE
• /
• v.42 no.4
• /
• pp.475-482
• /
• 2015

Developing a high-performance Semantic Role Labeling (SRL) system for a domain requires manually annotated training data of large size in the same domain. However, such SRL training data of sufficient size is available only for a few domains. Performances of Korean SRL are degraded by almost 15% or more, when it is directly applied to another domain with relatively small training data. This paper proposes two techniques to minimize performance degradation in the domain transfer. First, a domain adaptation algorithm for Korean SRL is proposed which is based on the prior model that is one of domain adaptation paradigms. Secondly, we proposed to use simplified features related to morphological and syntactic tags, when using small-sized target domain data to suppress the problem of data sparseness. Other domain adaptation techniques were experimentally compared to our techniques in this paper, where news and Wikipedia were used as the sources and target domains, respectively. It was observed that the highest performance is achieved when our two techniques were applied together. In our system's performance, F1 score of 64.3% was considered to be 2.4~3.1% higher than the methods from other research.

• Journal of Korea Multimedia Society
• /
• v.19 no.2
• /
• pp.240-251
• /
• 2016

This paper presents a real-time indoor location and pose estimation method that utilizes simple artificial markers and image analysis techniques for the purpose of warehouse automation. The conventional indoor localization methods cannot work robustly in warehouses where severe environmental changes usually occur due to the movement of stocked goods. To overcome this problem, the proposed framework places artificial markers having different interior pattern on the predefined position of the warehouse floor. The proposed algorithm obtains marker candidate regions from a captured image by a simple binarization and labeling procedure. Then it extracts maker interior pattern information from each candidate region in order to decide whether the candidate region is a true marker or not. The extracted interior pattern information and the outer boundary of the marker are used to estimate location and heading angle of the localization system. Experimental results show that the proposed localization method can provide high performance which is almost equivalent to that of the conventional method using an expensive LIDAR sensor and AMCL algorithm.

• Journal of Korea Game Society
• /
• v.10 no.5
• /
• pp.103-113
• /
• 2010

In this paper, we present a practical method for generating facial animation by using an optical motion capture system. In our setup, we assumed a situation of capturing the body motion and the facial expression simultaneously, which degrades the quality of the captured marker data. To overcome this problem, we provide an integrated framework based on the local coordinate system of each marker for labeling the marker data, hole-filling and removing noises. We justify the method by applying it to generate a short animated film.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.18 no.1
• /
• pp.133-142
• /
• 1993

This paper prrsents shape matching of two-dimensional object. This shape matching is applied to two-dimensional simple c10sedcurves represented by polygons. A large number of shape matching procedures have proposed baseed on teh view that shape can be represented by a vector of numerical features, and that this representation can be matched using techniques from statical pattern recognition. The varieties of features that have been extracted from shapes and used to represent them are numerous. But all of these feature-based approches suffer from the shortcoming that the descriptor of a segment of a shape do not ordinarily bear any simple relations hip to the description for the entire shape. We solve the segment matching problem of shape matching, defined as the recognition of a piece of a shape as approximate match to a part of large shape, by using relaxation labeling technique.

• The Korean Journal of Community Living Science
• /
• v.26 no.1
• /
• pp.63-74
• /
• 2015

A food allergy is described an adverse immunological reaction to a food item. It is increasingly common problem among infants, children, teenagers, and adults worldwide. This study examines food allergy knowledge, attitudes, practices, and health consciousness among college students studying food and nutrition and childhood education. A total of 235 food and nutrition and childhood education college students participated in the survey. According to the results, 41.3% of the respondents were aware of legal obligations associated labeling food items for food allergy; 14.0% were diagnosed with food allergy by their doctor; and 10.2% knew about food allergy symptoms. Food and nutrition students were more knowledgeable than childhood education students. The mean for food allergy attitudes was 4.22, and the score for food and nutrition students was higher than that for childhood education students. The mean for food allergy behaviors was 2.16, and the score of food and nutrition students was higher than that of childhood education students. The importance of food allergens was significantly higher than performance. These results suggest that, to improve the management of food allergies in foodservice operations, education programs regarding food allergies should be provided food and nutrition and childhood education students.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.9 no.6
• /
• pp.2302-2316
• /
• 2015

We propose an efficient method of extracting targets within a region of interest in non-homogeneous infrared images by using a principal component analysis (PCA) plane and adaptive Gaussian kernel. Existing approaches for extracting targets have been limited to using only the intensity values of the pixels in a target region. However, it is difficult to extract the target regions effectively because the intensity values of the target region are mixed with the background intensity values. To overcome this problem, we propose a novel PCA based approach consisting of three steps. In the first step, we apply a PCA technique minimizing the total least-square errors of an IR image. In the second step, we generate a binary image that consists of pixels with higher values than the plane, and then calculate the second derivative of the sum of the square errors (SDSSE). In the final step, an iteration is performed until the convergence criteria is met, including the SDSSE, angle and labeling value. Therefore, a Gaussian kernel is weighted in addition to the PCA plane with the non-removed data from the previous step. Experimental results show that the proposed method achieves better segmentation performance than the existing method.

• Proceedings of the Korean Society of Applied Pharmacology
• /
• 1998.11a
• /
• pp.158-158
• /
• 1998

The non-mevalonate pathway is a newly discovered isoprenoid biosynthetic pathway in some bacteria, cyanobacteria, algae and plants. Because isoprenoid metabolites (ubiquinone, menaquinone, undecaprenol) are essential for bacterial growth, this pathway may represent a novel target for antibacterial agents. Antibiotics with a unique mechanism of action are needed to combat the risk of antibiotic resistance that is a current worldwide problem. In order to study this pathway as viable target, it was necessary to verify use of the pathway in our model system, the bacterium Bacillus subtilis. Incubation experiments with [6,6-$^2$H$_2$]-D-glucose and [l-$^2$H$_3$]-deoxy-D-xylulose were conducted to provide labeled menaquinone-7 (MK -7), the most abundant isoprenoid in B. subtilis. $^2$H-NMR analysis of the MK-7 revealed labeling patterns that strongly support utilization of the non-mevalonate pathway. Another approach to study the pathway is by structure activity relationships of proposed inhibitors of the pathway. Fosmidomycin is a phosphonic acid with antibacterial activity known to inhibit isoprenoid biosynthesis in susceptible bacteria and may act by inhibiting the non-mevalonate pathway. Fosmidomycin and an N-methyl analog were synthesized and tested for antibacterial activity. Fosmidomycin was active against Escherichia coli and B. subtilis, while N-formyl-N-methyl-3-amino-propylphosphonic acid was inactive.

• Journal of Families and Better Life
• /
• v.22 no.1
• /
• pp.55-64
• /
• 2004

The major purpose of this study was to investigate purchasing behavior of environmental products by Korean consumers, which will ultimately help foster sustainable consumption. Environmental consciousness, environmental behaviors, level of awareness of environmental products, and purchasing of environmental products were examined. Mean differences between purchasers and non-purchasers of environmental products were compared in terms of environmental consciousness and behaviors, and the level of awareness of environmental products. A survey was conducted on 310 consumers in the greater Seoul metropolitan area. The data were analyzed by frequencies, percentages, logistic regression, and t-tests using a variable for interval scale and a variable for nominal scale. There were significant mean differences between purchasers and non-purchasers of environmental products on three variables of environmental consciousness and behaviors. Those who were educated on environmental issues showed a higher preference in purchasing environmental products. Among socio-demographic variables, the income level was the only variable that showed a significant mean difference between the two groups. Also, there was a remarkable difference in purchasing behavior between the two groups. For the purchasers of environmental products, the purchasing decisions took into account environment-friendliness of products. Non-purchasers of environmental products put more emphasis on price or quality of products. The results of the logistic regression analysis indicated that those who had higher education, who viewed environmental pollution as a serious problem, and who are more cognizant of the environmental labeling tend to purchase more environmental products.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.17 no.3
• /
• pp.129-138
• /
• 2022

Collecting and labeling sufficient training data, which is essential to deep learning-based visual inspection, is difficult for manufacturers to perform because it is very expensive. This paper presents a steel plate surface defect detection system with industrial-grade detection performance by training a small amount of steel plate surface images consisting of labeled and non-labeled data. To overcome the problem of lack of training data, we propose two data augmentation techniques: program-based augmentation, which generates defect images in a geometric way, and generative model-based augmentation, which learns the distribution of labeled data. We also propose a 4-step semi-supervised learning using pseudo labels and consistency training with fixed-size augmentation in order to utilize unlabeled data for training. The proposed technique obtained about 99% defect detection performance for four defect types by using 100 real images including labeled and unlabeled data.