• 문화기술의 융합
• /
• 제10권3호
• /
• pp.909-914
• /
• 2024

본 연구에서는 로봇 비전용 영상 인식을 비롯한 다양한 AI 분야에서 널리 활용되는 전이학습에 대한 정량적 평가를 제시하였다. 전이학습을 적용한 연구 결과에 대한 정량적, 정성적 분석은 제시되나, 전이학습 자체에 대해서는 논의되지 않는다. 따라서 본 연구에서는 전이학습 자체에 대한 정량적 평가를 숫자 손글씨 데이터베이스인 MNIST를 기반으로 제안한다. 기준 네트워크를 대상으로 전이학습 동결층의 깊이 및 전이학습 데이터와 사전 학습 데이터의 비율에 따른 정확도 변화를 추적하였다. 이를 통해 첫번째 레이어까지 동결할 때 전이학습 데이터의 비율이 3% 이상일 경우, 90% 이상의 정확도를 안정적으로 유지할 수 있음이 확인되었다. 본 연구의 전이학습 정량 평가 방법은 향후 네트워크 구조와 데이터의 종류에 따라 최적화된 전이학습을 구현하는데 활용 가능하며, 다양한 환경에서 로봇 비전 및 이미지 분석 AI의 활용 범위를 확대할 것이다.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2005년도 ICCAS
• /
• pp.751-755
• /
• 2005

We propose a sound source localization method using the Head-Related-Transfer-Function (HRTF) to be implemented in a robot platform. In conventional localization methods, the location of a sound source is estimated from the time delays of wave fronts arriving in each microphone standing in an array formation in free-field. In case of a human head this corresponds to Interaural-Time-Delay (ITD) which is simply the time delay of incoming sound waves between the two ears. Although ITD is an excellent sound cue in stimulating a lateral perception on the horizontal plane, confusion is often raised when tracking the sound location from ITD alone because each sound source and its mirror image about the interaural axis share the same ITD. On the other hand, HRTFs associated with a dummy head microphone system or a robot platform with several microphones contain not only the information regarding proper time delays but also phase and magnitude distortions due to diffraction and scattering by the shading object such as the head and body of the platform. As a result, a set of HRTFs for any given platform provides a substantial amount of information as to the whereabouts of the source once proper analysis can be performed. In this study, we introduce new phase and magnitude criteria to be satisfied by a set of output signals from the microphones in order to find the sound source location in accordance with the HRTF database empirically obtained in an anechoic chamber with the given platform. The suggested method is verified through an experiment in a household environment and compared against the conventional method in performance.

• Journal of Radiation Protection and Research
• /
• 제28권2호
• /
• pp.145-154
• /
• 2003

By optimizing the radiation protection the collective dose and individual dose could be reduced during YGN #4 $5^{th}$ outage in 2001. The collective doses for the two high radiation jobs decreased to 85% and 65% of expected doses. The proportion of workers with low dose (below 1mSv) exposure increased 4% while the proportion of workers with over 3mSv and 5mSv exposure are decreased to 2%, 1% respectively. But none is exposed over 8mSv for the annual dose. To aid decision of utilizing the robot, cost- benefit analysis was performed and reasonable point was proposed to use the robot. For the first time job, repeated ALARA meeting and mock up training were implemented to set up working procedure by identifying the trouble. To easily set up standard procedure, mockup process was videotaped and reviewed during ALARA meeting. Monitoring is a good approach to chase radiological working condition such as working time, dose rate. behavior of workers, especially for high radiation work. Those data were estimated and adjusted from the stage of work planning to mock up. At the stage of actual work the monitoring data were compared to the estimation and recorded to database. This database will not only be used as a powerful tool for dose optimization at the following outage but also as a guideline to dose constraint set up for optimization for each specific situation.

• 대한전자공학회논문지
• /
• 제22권1호
• /
• pp.53-60
• /
• 1985

A computer vision system for robot is developed which can recognize a variety of two dimensional complex objects in gray level noisy scenes. the system is also capable of determining the position and orientation of the objects for robotlc manipulation. The hardware of the vision system is developed and a new edge tracking technique is also proposed. The linked edges are approximated to sample line drawing by split and merge algorithm. The system extracts many features from line drawing and constructs relational structure by the concave and convex hull of objects. In matching process, the input obhects are compared with the objects database which is formed by learning ability. Thelearning process is so simple that the system is very flexible. Several examples arc shown to demonstrate the usefulness of this system.

• 제어로봇시스템학회논문지
• /
• 제11권1호
• /
• pp.50-57
• /
• 2005

We present a simple and effective method of face and facial feature detection under pose variation of user face in complex background for the human-robot interaction. Our approach is a flexible method that can be performed in both color and gray facial image and is also feasible for detecting facial features in quasi real-time. Based on the characteristics of the intensity of neighborhood area of facial features, new directional template for facial feature is defined. From applying this template to input facial image, novel edge-like blob map (EBM) with multiple intensity strengths is constructed. Regardless of color information of input image, using this map and conditions for facial characteristics, we show that the locations of face and its features - i.e., two eyes and a mouth-can be successfully estimated. Without the information of facial area boundary, final candidate face region is determined by both obtained locations of facial features and weighted correlation values with standard facial templates. Experimental results from many color images and well-known gray level face database images authorize the usefulness of proposed algorithm.

• 한국정밀공학회지
• /
• 제19권10호
• /
• pp.66-76
• /
• 2002

Automated welding and soldering are an important manufacturing issue in order to lower the cost, increase the quality, and avoid labor problems. An off-line programming, OLP, is one of the powerful methods to solve this kind of diversity problem. Unless an OLP system is ready for the path optimization in welding and soldering, the waste of time and cost is unavoidable due to inefficient paths in welding and soldering processes. Therefore, this study attempts to obtain path optimization using a genetic algorithm based on artificial intelligences. The problem of welding path optimization is defined as a conventional TSP (traveling salesman problem), but still paths have to go through welding lines. An improved genetic algorithm was suggested and the problem was formulated as a TSP problem considering the both end points of each welding line read from database files, and then the transit problem of welding line was solved using the improved suggested genetic algorithm.

• 제어로봇시스템학회논문지
• /
• 제16권3호
• /
• pp.277-285
• /
• 2010

This paper describes an approach to analyze geometrical information of building images for understanding outdoor environment of autonomous navigation robot. Line segments and color information are used to classily a building with the other objects such as sky, trees, and roads. The line segments and their two neighboring regions are extracted from detected edges in image. The model of line segment (MLS) consists of color information of neighbor regions. This model rules out the line segments of non-building face. A building face converges into dominant vanishing points (DVPs) which include one vertical point and one of five horizontal points in maximum. The intersection of vertical and horizontal lines creates a facet of building. The geometrical characteristics such as the center coordinates, area, aspect ratio and aligned coexistence are used for extracting the windows in the building facet. In experiments, 150 building faces and 1607 windows were detected from the database of outdoor environment. We found that this result shows 94.46% detection rate. These experimental images were all taken in Ulsan metropolitan city in Korea under difference of viewpoints, daytime, camera system and weather condition.

• 한국지능시스템학회논문지
• /
• 제16권5호
• /
• pp.587-593
• /
• 2006

영상을 통한 감정 인식 기술은 사회의 여러 분야에서 필요성이 대두되고 있음에도 불구하고 인식 과정의 어려움으로 인해 풀리지 않는 문제로 남아 있다. 인간의 움직임을 이용한 감정 인식 기술은 많은 응용이 가능하기 때문에 개발의 필요성이 증대되고 있다. 영상을 통해 감정을 인식하는 시스템은 매우 다양한 기법들이 사용되는 복합적인 시스템이다. 따라서 이를 설계하기 위해서는 영상에서의 움직임 추출, 특징 벡터 추출 및 패턴 인식 등 다양한 기법의 연구가 필요하다. 본 논문에는 이전에 연구된 움직임 추출 방법들을 바탕으로 한 새로운 감정 인식 시스템을 제안한다. 제안된 시스템은 은닉 마르코프 모델을 통해 동정된 분류기를 이용하여 감정을 인식한다. 제안된 시스템의 성능을 평가하기 위해 평가데이터 베이스가 구축되었으며, 이를 통해 제안된 감정 인식 시스템의 성능을 확인하였다.

• 한국차세대컴퓨팅학회논문지
• /
• 제15권3호
• /
• pp.16-24
• /
• 2019

최근 뇌파를 기반으로 한 인간의 감정을 인식하는 연구가 인간-로봇 상호작용 분야에서 활발히 진행되고 있다. 본 논문에서는 MAHNOB-HCI에서 사용된 자기평가와 주석 레이블링 방법과는 다른, 이미지 기반의 뇌파 Topography를 이용한 레이블링을 통해 감정을 평가하는 방법을 제안한다. 제안한 방법은 뇌파 신호를 Topography의 이미지로 변환하여 기계학습 모델을 학습하고 이를 기반으로 Valence 기반의 감정을 평가한다. 제안한 방법은 레이블링 과정을 자동화하여 지연 시간을 없애고 객관적인 레이블링을 제공할 수 있다. MAHNOB-HCI 데이터베이스를 적용한 실험에서 SVM, kNN의 기계학습 모델을 학습하여 주석 레이블링과 성능 비교를 하였으며, 제안 방법의 감정인식 정확도를 SVM에서 54.2%, kNN에서 57.7%로 확인하였다.

• 로봇학회논문지
• /
• 제19권2호
• /
• pp.221-228
• /
• 2024

As the importance of cardiovascular health is highlighted, research on its correlation with blood pressure, the most important indicator, is being actively conducted. Therefore, extensive clinical data is essential, but the measurement of the central arterial blood pressure waveform must be performed invasively within the artery, so the quantity and quality are limited. This study suggested a mock circulatory robot and artificial aorta to reproduce the blood pressure waveform generated by the overlap of forward and reflected waves. The artificial aorta was fabricated with biomimetic silicone to mimic the physiological structure and vascular stiffness of the human. A pressurizing chamber was implemented to prevent distortion of the blood pressure waveform due to the strain-softening of biomimetic silicone. The reproduced central arterial blood pressure waveforms have similar magnitude, shape, and propagation characteristics to humans. In addition, changes in blood pressure waveform due to aging were also reproduced by replacing an artificial aorta with various stiffness. It can be expanded to construct a biosignal database and health sensor testing platform, a core technology for cardiovascular health-related research.