• 한국재난정보학회 논문집
• /
• 제18권4호
• /
• pp.883-890
• /
• 2022

연구목적:현재 국내 소방시설설계의 경우 낮은 설계단가와 업체 간 과열 경쟁으로 고급 인력에 대한 확보가 어려워 건축물의 화재안전성능을 향상시키는데 한계가 있다. 이에 이러한 문제를 해소하고 선도적인 소방엔지니어링 기술을 확보하기 위해 AI 기반 소방설계솔루션을 연구하였다. 연구방법: 기존 소방설계에 많이 사용되는 AutoCAD를 통해 기본 설계 및 실시 설계에 필요한 절차를 프로세스화 하고 YOLO v4 객체 인식 딥러닝 모델을 통해 AI기술을 활용하였다. 연구결과:소방시설에 대한 설계프로세스를 통해 설비의 결정과 도면 설계 자동화를 진행하였다. 또한 문 및 기둥에 대한 이미지를 학습시켜 인공지능이 해당 부분을 인식하여 경계구역 선정, 배관 및 소방시설을 설치하는 기능을 구현하였다. 결론:인공지능 기술을 기반으로 건축물 화재방호 설비에 대한 기본 및 실시 설계 도면 작성 시 인적 및 물적 자원을 저감시킬 수 있을 것으로 확인되었으며 선행적인 기술 개발을 통해 인공지능 기반 소방설계에 기술력을 확보하였다.

• Asian Pacific Journal of Cancer Prevention
• /
• 제16권16호
• /
• pp.7117-7121
• /
• 2015

Background: The effects of home nursing intervention on the quality of life in patients with nasopharyngeal carcinoma (NPC) after radiotherapy and chemotherapy are unclear. According to the characteristics of nursing home patients with nasopharyngeal carcinoma, we should continuously improve the nursing plan and improve the quality of life of patients at home. Materials and Methods: We selected 180 patients at home with NPC after radiotherapy and chemotherapy. The patients were randomly divided into experimental and control groups (90 patients each). The experimental group featured intervention with an NPC home nursing plan, while the control group was given routine discharge and outpatient review. Nursing intervention for patients was mainly achieved by regular telephone follow-up and home visits. We use the quality of life scale (QOL-C30), anxiety scale (SAS) and depression scale (SDS) to evaluate these patients before intervention, and during follow-up at 1 month and 3 months after the intervention. Results: Overall health and quality of life were significantly different between the groups (p<0.05), Emotional function score was significantly higher after intervention (p<0.05), as were cognitive function and social function scores after 3 months of intervention (p<0.05). Scores of fatigue, nausea and vomiting, pain, appetite and constipation were also significantly different between the two groups (p<0.05). Rates of anxiety and depression after 3 months of intervention were 11.1%, 22.2% and 34.4%, 53.3%, the differences being significant (p<0.05). Conclusions: NPC home nursing plan could effectively improve overall quality of life, cognitive function, social function (after 3 months) of patients, but improvement regarding body function is not suggested. Fatigue, nausea and vomiting, pain, appetite, constipation were clearly improved. We should further pursue a personalized, comprehensive measurements for nursing interventions and try to improve the quality of life of NPC patients at home.

• International Journal of Advanced Culture Technology
• /
• 제6권4호
• /
• pp.262-265
• /
• 2018

Deep learning, a sub-field of machine learning changing the prospects of artificial intelligence (AI) because of its recent advancements and application in various field. Deep learning deals with algorithms inspired by the structure and function of the brain called artificial neural networks. This works reviews basic architecture and recent advancement of deep structured learning. It also describes contemporary applications of deep structured learning and its advantages over the treditional learning in artificial interlligence. This study is useful for the general readers and students who are in the early stage of deep learning studies.

• 정보보호학회논문지
• /
• 제33권5호
• /
• pp.761-773
• /
• 2023

본 논문에서는 딥러닝 분야에서 사용되는 신경망 모델, 그중에서도 다중 계층 퍼셉트론 모델에 사용되는 지수함수 기반의 활성화 함수를 근사 함수로 대체하고, 근사 함수에 마스킹을 적용함으로써 신경망 모델의 추론 과정의 전력 분석 저항성을 높이는 방법을 제안한다. 이미 학습된 값을 사용하여 연산하는 인공 신경망의 추론 과정은 그 특성상 가중치나 편향 등의 내부 정보가 부채널 공격에 노출될 위험성이 있다. 다만 신경망 모델의 활성화 함수 계층에서는 매우 다양한 함수를 사용하고, 특히 지수함수 기반의 활성화 함수에는 마스킹 기법 등 통상적인 부채널 대응기법을 적용하기가 어렵다. 따라서 본 연구에서는 지수함수 기반의 활성화 함수를 단순한 형태로 근사하여도 모델의 치명적인 성능 저하가 일어나지 않음을 보이고, 근사 함수에 마스킹을 적용함으로써 전력 분석으로부터 안전한 순방향 신경망 모델을 제안하고자 한다.

• 한국게임학회 논문지
• /
• 제21권3호
• /
• pp.107-116
• /
• 2021

바둑에서의 형세 판단은 현재 대국 중인 흑백 대국자 간의 유불리를 판단하는 척도가 되며, 이를 통해 곧바로 적절한 전술과 전략을 구사하게 된다. 본 논문에서는 거리에 따라 반감하는 세력 함수를 활용하여 알파고 간의 50개 대국의 형세 판단을 하고자 했다. 실험 결과에 따르면 단지 세력 함수만을 사용하여 형세 판단을 하게 되면 정확한 판단을 함에 한계가 있음이 밝혀졌다. 이를 극복하기 위해 사석 처리를 위한 사활문제 해결이 필요하며, 이를 보강하게 되면 바둑에서의 정밀한 형세 판단을 할 수 있음을 보였다.

• International journal of advanced smart convergence
• /
• 제11권3호
• /
• pp.132-141
• /
• 2022

We look at the rocket life stages of NaverZ's metaverse platform scaling and investigate the ignition and scale-up stage of its metaverse platform brand, Zepeto based on the Rocket Model (RM). The results are derived as follows: Firstly, NaverZ shows the event strategy by collaborating with K-pops, the piggybacking strategy by utilizing other SNSs, and the VIP strategy by investing in game and entertainment content genres in the 'attract' function. In the second 'match' function, based on the matching rule of Zepeto, the users can generate their own characters and "World" with Zepeto Studio. However, for strengthening the matching quality, NaverZ is investing in the artificial intelligence (AI) based companies consistently. In the 'connect' function, NaverZ's maximization of the positive interaction is possible by inducing feed activities in Zepeto & other SNSs and by uploading attractive content for viral effects in the ignition. For facilitating this, NaverZ expands the scale to other continents like Southeast Asia and Middle East with the localization strategy inclusive investment. Lastly, in the 'transact' function, based on three monetization experiments like Coin & ZEM, user generated content (UGC) fee, and advertising revenue in the ignition, NaverZ starts to invest in NFT platforms and abroad blockchain companies.

• 한국의류산업학회지
• /
• 제25권6호
• /
• pp.739-744
• /
• 2023

With the popularization of 5G networks and the development of AI (artificial intelligence) technology, Metaverse, which creates production capacity by combining virtual space and reality, is attracting attention. In this study, we searched for makeup applications with more than 100 million downloads from October 11, 2020 to November 3, 2020 through the Google Play Store. As a result of the search, four applications were found: YouCam Makeup, YouCam Perfect, Beauty Plus, and Sweet Snap. Based on the functions provided by the four applications, we attempted to suggest makeup functions applicable to Zepeto's avatar. Functions for the eyes (eyeliner, eyelashes, mascara, eye shadow, eye shape, eyebrow shape, lenses, double eyelids), functions for the nose (nose shape), functions for the mouth (lipstick, lip shape, smile function) ) Functions corresponding to the facial contour (contour, skin foundation, blusher, shading, highlighter, face painting, theme makeup) and functions corresponding to the body (body adjustment) were proposed. This study is the first in the beauty field to propose a method of applying the functions of the Metaverse platform as the importance of digital platforms is highlighted, and is the first to propose a makeup function applied to the Metaverse so that it can be used as important basic data in the future.

• 한국항공운항학회지
• /
• 제18권2호
• /
• pp.16-22
• /
• 2010

This paper proposes the real-time moving object tracking system UAV using color information. Case of object tracking, it have studied to recognizing the moving object or moving multiple objects on the fixed camera. And it has recognized the object in the complex background environment. But, this paper implements the moving object tracking system using the pan/tilt function of the camera after the object's region extraction. To do this tracking system, firstly, it detects the moving object of RGB/HSI color model and obtains the object coordination in acquired image using the compact boundary box. Secondly, the camera origin coordination aligns to object's top&left coordination in compact boundary box. And it tracks the moving object using the pan/tilt function of camera. It is implemented by the Labview 8.6 and NI Vision Builder AI of National Instrument co. It shows the good performance of camera trace in laboratory environment.

• 로봇학회논문지
• /
• 제15권4호
• /
• pp.398-403
• /
• 2020

Robot's throwing control is difficult to accurately calculate because of air resistance and rotational inertia, etc. This complexity can be solved by using machine learning. Reinforcement learning using reward function puts limit on adapting to new environment for robots. Therefore, this paper applied deep reinforcement learning using neural network without reward function. Throwing is evaluated as a success or failure. AI network learns by taking the target position and control policy as input and yielding the evaluation as output. Then, the task is carried out by predicting the success probability according to the target location and control policy and searching the policy with the highest probability. Repeating this task can result in performance improvements as data accumulates. And this model can even predict tasks that were not previously attempted which means it is an universally applicable learning model for any new environment. According to the data results from 520 experiments, this learning model guarantees 75% success rate.

• 한국전자통신학회논문지
• /
• 제15권2호
• /
• pp.299-306
• /
• 2020

신경망을 이용한 영상인식은 여러 분야에 널리 사용되고 있다. 본 연구에서는 차량 번호 인식 및 특정 구역 입출 시 통제에 필요한 인가/비인가 차량 인식 시스템을 연구하였다. 이 시스템은 영상을 인식하는 기능을 갖추고 있어 차량 번호에 대한 모든 정보를 확인하고, 차량 번호판을 정확히 인식할 수 있는 기능을 추가하였다. 그 밖에 신경망을 이용하여 좀 더 빠르게 차량번호를 확인할 수 있도록 하였다.