• International Journal of Computer Science & Network Security
• /
• v.22 no.3
• /
• pp.273-275
• /
• 2022

Deep Learning is used nowadays in Nuclei segmentation. While recent developments in theory and open-source software have made these tools easier to implement, expert knowledge is still required to choose the exemplary model architecture and training setup. We compare two popular segmentation frameworks, U-Net and Mask-RCNN, in the nuclei segmentation task and find that they have different strengths and failures. we compared both models aiming for the best nuclei segmentation performance. Experimental Results of Nuclei Medical Images Segmentation using U-NET algorithm Outperform Mask R-CNN Algorithm.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2020.11a
• /
• pp.331-334
• /
• 2020

항만에서 액체화물 취급이 증가함에 따라 로딩암의 사용빈도가 늘어나고 있다. 이에 본 논문에서는 로딩암 교육생들이 실전에 투입되기 전에 충분히 교육할 수 있도록 교육용 시뮬레이션을 제시한다. 그와 더불어 시뮬레이션과 연동되는 로딩암모형의 디지털트윈 연동을 통해, 교육생들의 로딩암 조종기술 향상을 이루어, 보다 안전한 로딩암작업과 항만의 액체화물사고 빈도를 크게 줄일 것으로 기대한다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2023.07a
• /
• pp.29-32
• /
• 2023

본 논문에서는 외국인과 발음에 어려움을 겪고 있는 한국인들을 대상으로 음성 녹음을 진행하여 점수를 반환받는 웹 사이트를 소개한다. 이 웹 사이트의 목적은 사용자들의 발음 향상을 돕는 것이다. 음성 인식 API와 발음 평가 API를 사용하여 사용자의 발음을 정확하게 평가하고 피드백을 제공함으로써, 외국어 학습자와 발음에 어려움을 겪는 한국인들이 보다 원활하게 의사소통할 수 있도록 돕는다. 향후 연구로는 이 시스템의 사용자들에게 학습 성취에 대한 동기 부여를 하는 기능을 추가해 학습 효과를 높이도록 개선할 것이다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2008.11a
• /
• pp.211-214
• /
• 2008

Brain Computer Interface(BCI)분야는 뇌파를 이용하여 컴퓨터를 컨트롤 하는 기술로 최근 많은 연구가 이루어 지고 있다. 뇌파는 주변 상황과 개인, 상태에 따라 그 변화가 명확하기 때문에 BCI 분야는 앞으로 많은 응용 프로그램 개발에 충분한 자원이 될 수 있다. 기존의 BCI 연구는 뇌파를 입력 값으로 사용하여 컴퓨터를 컨트롤 하였다. 하지만 뇌파 값은 환경과 상황, 개인마다 다르기 때문에 특정 값으로 사용하기에 어려운 점이 있다. 본 논문에서는 이러한 뇌파의 특징을 이용하여 집중력을 향상시키는 개인용 게임시스템을 제안하고자 한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2008.11a
• /
• pp.1462-1464
• /
• 2008

본 논문에서는 정보시스템 운영시 발생하는 사용자 취약성(Human Vulnerability) 문제의 심각성에 대해 알아보고 이를 개선하기 위한 교육 및 훈련 프로그램을 다루고 있는 기존 관련 연구들을 조사 분석 한다. 아울러 기존 연구에서 보완되어야 할 개선 요구사항 들을 도출하여 향후 효과적인 취약성 개선 프로그램 제공을 위한 가상머신에 기반한 보안 훈련장 시스템 아키텍쳐를 제안한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2024.05a
• /
• pp.741-744
• /
• 2024

본 연구에서는 음성 전처리 기법인 푸리에 변환의 높은 시간 복잡도로 인해 많은 계산 자원을 요구한다는 단점을 보완하기 위한 FTAE(Fourier Transform Auto Encoder)를 설계하고 구현한다. FTAE는 음성 데이터를 입력으로 받아 Early Fusion 특징맵을 출력하도록 설계된 오토인코더 기반 신경망이다. 학습 결과 FTAE의 최종 Training Loss는 0.1479를 나타냈다. 기존 푸리에 변환 기반 Early Fusion 방법과의 성능 비교 실험 결과 FTAE 방법은 Accuracy 0.905, F1-Score 0.905, 탐지 소요 시간 17초의 성능을 보였다. FTAE 방법은 Early Fusion 방법에 비해 Accuracy와 F1-Score는 0.065 하락했지만, 탐지 소요 시간은 약 72배 빠른 결과를 보여주었다.

• Journal of The Korean Association of Information Education
• /
• v.21 no.2
• /
• pp.171-181
• /
• 2017

Each City & Provincial office of education has provided educational informatization training, which invites its partner country's teachers as part of 'e-Learning Globalization Project.' The purposes of this training program are, a) improving nation brand power by maintaining e-learning national competitiveness and strengthening influence, b) solving digital divide by sharing Korean best practices, c) establishing partnership through exchanges and cooperation, d) capacity building which support partner country's leading teachers can play key roles to spread educational informatization in their own country. It is necessary to reflect outcomes of training program until now and to find sustainable development plan, at 10th anniversary of Education ICT training program now. In order to fulfil purposes, qualitative research was conducted to analyze current status based on context, input, process, and product phases, and to find sustainable development alternative. In-depth interview(IDI) was conducted among supervisor, officials, and teachers from partner countries. Also, the result of IDI was analyzed by themes.

• Journal of Information Processing Systems
• /
• v.8 no.4
• /
• pp.575-588
• /
• 2012

Exploiting unlabeled text data with a relatively small labeled corpus has been an active and challenging research topic in text mining, due to the recent growth of the amount of biomedical literature. Biomedical named-entity recognition is an essential prerequisite task before effective text mining of biomedical literature can begin. This paper proposes an Active Co-Training (ACT) algorithm for biomedical named-entity recognition. ACT is a semi-supervised learning method in which two classifiers based on two different feature sets iteratively learn from informative examples that have been queried from the unlabeled data. We design a new classification problem to measure the informativeness of an example in unlabeled data. In this classification problem, the examples are classified based on a joint view of a feature set to be informative/non-informative to both classifiers. To form the training data for the classification problem, we adopt a query-by-committee method. Therefore, in the ACT, both classifiers are considered to be one committee, which is used on the labeled data to give the informativeness label to each example. The ACT method outperforms the traditional co-training algorithm in terms of f-measure as well as the number of training iterations performed to build a good classification model. The proposed method tends to efficiently exploit a large amount of unlabeled data by selecting a small number of examples having not only useful information but also a comprehensive pattern.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.11 no.11
• /
• pp.4244-4250
• /
• 2010

This study analyzed relation with accounting information system operations of more integration and end-user's business processing capabilities. Also, Studied the correlation that has reflecting between business computer skills and training experience level of end-user in practical operation. In summary of this study, first, found out that the improvement and training level of computerized accounting training in mid-size enterprise can be processing more effectively an accounting information business. Second, the end-user training and business development of business processes using the system, the positive impact was significant. Finally, the introduction of the continuous integration system based on new technology analyzed more lower overall performance satisfaction of end-user in the tasks. Business Ability of end-user could be important factor as ongoing business development and training. The results of this study, It can support to effective strategies of end-user systems, will expected to further strengthen competitiveness in the future.

• Journal of the Korea Convergence Society
• /
• v.12 no.7
• /
• pp.263-271
• /
• 2021

Optoelectronic shooting training systems are used in shooting training sites to improve the accuracy of shooting by tracking the trajectories of gun movements. However, optoelectronic-based systems have limitations in terms of cost, complexity of installation, and the risk that electronic targets may be broken. In this study, we developed and verified a shooting training system that measures postural tremors using a low-cost accelerometer. The acceleration sensor module was designed to be attached to the air cylinder of a gun. Postural tremors were evaluated based on amplitude, frequency, and spatial pattern index, which were computed using acceleration data. The postural tremor indices between the accelerometer and optoelectronic-based system were highly correlated (left-right and up-down directions: r = 0.76 and r = 0.70, respectively). We validated the developed shooting training system using an independent two-sample t-test, which identified a significant difference (p < 0.05) in the calculated postural tremor index according to the athlete's shooting score (i.e., best and worst shots).