• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.24 no.2
• /
• pp.87-92
• /
• 2024

In this paper, a RAG system was implemented using an existing Large Language Model (LLM) and Langchain library to develop a dental consultation chatbot. For this purpose, we collected contents from the webpage bulletin boards of domestic dental university hospitals and constructed consultation data with the advice and supervision of dental specialists. In order to divide the input consultation data into appropriate sizes, the chunk size and the size of the overlapping text in each chunk were set to 1001 and 100, respectively. As a result of the simulation, the Retrieval Augmented LLM searched for and output the consultation content that was most similar to the user input. It was confirmed that the accessibility of dental consultation and the accuracy of consultation content could be improved through the built chatbot.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2006.11a
• /
• pp.273-276
• /
• 2006

본 논문에서는 일반적인 수족관에서 물고기를 관찰하는데 있어서 관찰자가 빠르고 효율적으로 물고기에 대한 정보를 얻을 수 있는 방법을 제안하고자 한다. 여러 종의 물고기가 살고 있는 수족관 환경에서 처음 보는 관찰자과 쉽게 물고기에 대한 정보를 얻기는 힘들다. 이러한 제약을 효과적으로 개선하기 위하여 반 투영 거울을 사용하는 공간 증강 현실을 위한 영상출력 기법을 이용한다. 이때 피사체의 위치 정보를 얻기 위해 배경인 수족관영상과 피사체인 물고기영상을 분리한 후 물고기의 깊이 정보 값을 추출하기 위하여 호모그래피 행렬을 이용하고 보정된 이미지를 계산 한다. 실험결과 본 논문에서 제안한 방법을 통해 구축한 이 시스템으로 관찰자는 각각의 물고기들에 대한 정보를 반응형 Annotation으로 쉽게 얻을 수 있었다.

• Annual Conference on Human and Language Technology
• /
• 2023.10a
• /
• pp.346-351
• /
• 2023

오픈 도메인 대화 시스템은 산업에서 다양하게 활용될 수 있지만 유해한 응답을 출력할 수 있다는 위험성이 지적되어 왔다. 본 논문에서는 언급된 위험성을 완화하기 위해 데이터 측면에서 대화 시스템 모델을 개선하는 방법을 제안한다. 대화 모델의 유해한 응답을 유도하도록 설계된 데이터셋을 사용하여 모델이 올바르지 못한 응답을 생성하게 만들고, 이를 LLM을 활용하여 안전한 응답으로 수정한다. 또한 LLM이 정확하게 수정하지 못하는 경우를 고려하여 추가적인 필터링 작업으로 데이터셋을 보완한다. 생성된 데이터셋으로 추가 학습된 대화 모델은 기존 대화 모델에 비해 대화 일관성 및 유해성 면에서 성능이 향상되었음을 확인했다.

• Journal of the Korea Convergence Society
• /
• v.11 no.10
• /
• pp.81-88
• /
• 2020

This purpose of this study is to develop content that enables repetitive carving practice of the maxillary right central incisor (MRCI) based on augmented reality (AR). For a step-by-step practice of achieving the tooth shape, after creation of the storyboard from the square box shape in step 1 to the completed MRCI block in step 16, three-dimensional (3D) modeling data reflecting the characteristics of the mesial, distal, lingual, and labial surface of the MRCI were generated. An application was built in which 3D modeling data were output on the screen of the learner's mobile device, and image markers suitable for 3D modeling in steps 1 to 16 of the MRCI model were respectively generated. Using this information, the learner could carve a high-quality MRCI by repeatedly performing the tooth shape carving exercises. With AR, we intend to contribute to improved tooth morphology carving skills by linking the theory and practical techniques for a beginners in dentistry.

• Journal of agriculture & life science
• /
• v.51 no.1
• /
• pp.205-212
• /
• 2017

This study aimed to output augmented reality display using chroma-key so that power tiller simulator can be operated smoothly while wearing the head mounted display. In this paper, we propose a chroma-eliminating image filtering method. Experimental results show that the maximum and minimum values of hue, saturation and intensity were 0.52, 0.153, 0.57, 0.16, 1, and 0.12, respectively. A keypad was used to set the initial position of augmented reality adjusted with the front, back, top, bottom, left, and right buttons. The initial position value is always maintained and managed according to the trailer attachment and detachment. Finally, we show that the augmented reality merged with virtual image and the acquired image of operation device using coordinate values obtained from the HMD and the position tracking sensor as relative coordinates in Unity program that is the ultimate game development platform.

• KIPS Transactions on Computer and Communication Systems
• /
• v.9 no.12
• /
• pp.333-338
• /
• 2020

This paper proposes an augmented reality-based electronic circuit education system as a way for electronic circuit education, which is the basis of ICT convergence technology field. It consists of a hardware module that can identify the actual circuit and a mobile educational content that can check the current flow, input, output, and measured value by applying augmented reality technology. An experiment was conducted on image recognition, which is the main performance, for the purpose of stable operation of the system, and as the experimental method the recognition rate was measured by changing the distance between the hardware module and the mobile device to a certain interval. As a result of the experiment, the recognition rate was 100 percent at a distance of 25[Cm] or higher, and it was confirmed that the recognition rate decreased by 12% at a distance below 25[Cm], which can be said to be the effect of an error that results in image loss taken due to close distance. In the future, we plan to apply the education system presented in this paper to classes, which increases the efficiency of classes and improve students' interest and understanding of the subject.

• Journal of the Institute of Convergence Signal Processing
• /
• v.12 no.4
• /
• pp.258-266
• /
• 2011

AR(Augmented Reality) technology is now easily shown around us with respect to its applicable areas' being spreaded into various shapes since the usage is simply generalized and many-sided. Currently existing camera vision based AR used marker based methods rather than using real world's informations. For the marker based AR technology, there are limitations on applicable areas and its environmental properties that a user could immerse into the usage of application program. In this paper, we proposed a novel AR method which users could recognize objects from the real world's data and the related 3-dimensional contents are also displayed. Those are done using image processing skills and a smart mobile embedded camera for terminal based AR implementations without any markers. Object recognition is done from the comparison of pre-registered and referenced images. In this process, we tried to minimize the amount of computations of similarity measurements for improving working speed by considering features of smart mobile devices. Additionally, the proposed method is designed to perform reciprocal interactions through touch events using smart mobile devices after the 3-dimensional contents are displayed on the screen. Since then, a user is able to acquire object related informations through a web browser with respect to the user's choice. With the system described in this paper, we analyzed and compared a degree of object recognition, working speed, recognition error for functional differences to the existing AR technologies. The experimental results are presented and verified in smart mobile environments to be considered as an alternate and appropriate AR technology.

• Corrosion Science and Technology
• /
• v.15 no.3
• /
• pp.135-140
• /
• 2016

Piping and equipment are degraded by flow-accelerated corrosion (FAC) in nuclear power plants. FAC causes numerous problems and nuclear utilities maintain programs to control FAC. The key parameters influencing FAC are hydrodynamic conditions, water chemistry, and effect of materials. Recently, a nuclear utility has planned slight power uprates in Korea. Operating conditions need to be changed in the secondary system according to power uprates. This study analyzed the effect of wall-thinning caused by power uprates. The change of operation data in the secondary cycle is reviewed, and wall-thinning rates are analyzed in the main lines. As a result, two phase (mixture of water and steam) lines have a greater impact than a water line under power uprate conditions. Also, the quality of steam is the most important factor for FAC in two phase lines.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2022.06a
• /
• pp.217-220
• /
• 2022

본 논문에서는 CCTV 영상 화질을 향상하고 해상도를 높이기 위해 딥 러닝(Deep Learning)을 이용하여 잡음 제거(Denoising) 와 초해상도(Super-resolution) 작업을 수행한다. 데이터 증강(Data Augmentation)을 통한 초해상도 성능 향상을 위해서 잡음 제거 네트워크의 출력 영상을 초해상도 네트워크의 입력으로 사용하는 순차적 작업을 사용한다. 또한 딥 러닝을 이용한 영상처리에서 발생하는 평균 밝기 오차 문제를 해결하기 위한 손실함수(Loss Function)와 두 가지 이상의 순차적인 딥 러닝 작업에서 발생하는 문제점을 극복하기 위한 손실함수를 제안한다. 제안하는 손실함수는 네트워크의 출력 영상과 타겟 영상의 밝기 오차를 줄이는 것이 가능하고, 순차적 작업에서 보다 정확한 모델 성능 판단이 가능하다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.11a
• /
• pp.800-802
• /
• 2022

최근 성인들을 위한 '키덜트 장난감'의 관심도가 증가하고 있다. 그중 높은 매출을 보이는 RC 카와 VR 의 콘텐츠 부족 및 대중화 문제를 해결하면 시장규모를 성장시킬 수 있다고 판단한다. 본 논문은 차별화된 RC 카 레이싱 게임 개발을 목표로 한다. 스마트폰용 VR 기기를 착용해 영상을 보며 컨트롤러로 RC 카를 조종한다. 또한 실시간 객체 검출이 가능한 YOLOv5 를 활용해 표지판 인식 및 바운딩 박스, 표지판 라벨, 라벨 음성 출력 기능과 오픈 소스 기반 실시간 컴퓨터 비전인 OpenCV 기반 알고리즘을 활용하여 차선을 인식해 이를 기반으로 영상 처리를 거쳐 가상 차선 및 가상 트랙을 출력한다. 결론적으로 RC 카와 VR 로 구현하여 이를 통해 부족한 VR 컨텐츠를 추가하고 접근성을 강화한다.