• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.20 no.6
• /
• pp.313-330
• /
• 2021

In an autonomous driving system equipped with a camera and lidar, depth completion techniques enable dense depth estimation. In particular, using self-supervised learning it is possible to train the depth completion network even without ground truth. In actual autonomous driving, such depth completion should have very short latency as it is the input of other algorithms. So, rather than complicate the network structure to increase the accuracy like previous studies, this paper focuses on network latency. We design a U-Net type network with RegNet encoders optimized for GPU computation. Instead, this paper presents several techniques that can increase accuracy during the process of self-supervised learning. The proposed techniques increase the robustness to unreliable lidar inputs. Also, they improve the depth quality for edge and sky regions based on the semantic information extracted in advance. Our experiments confirm that our model is very lightweight (2.42 ms at 1280x480) but resistant to noise and has qualities close to the latest studies.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.11
• /
• pp.155-160
• /
• 2023

This paper presents a method for predicting blood pressure using the photoplethysmography signals. First, after measuring the optical blood flow signal, artifacts are removed through a preprocessing process, and a signal for learning is obtained. In addition, weight and height, which affect blood pressure, are measured as additional information. Next, a system is built to estimate systolic and diastolic blood pressure by learning the photoplethysmography signals, height, and weight as input variables through an artificial intelligence algorithm. The constructed system predicts the systolic and diastolic blood pressures using the inputs. The proposed method can continuously predict blood pressure in real time by receiving photoplethysmography signals that reflect the state of the heart and blood vessels, and the height and weight of the subject in an unconstrained method. In order to confirm the usefulness of the artificial intelligence-based blood pressure prediction system presented in this study, the usefulness of the results is verified by comparing the measured blood pressure with the predicted blood pressure.

• Trans-
• /
• v.8
• /
• pp.95-127
• /
• 2020

This research is based on using Bartenieff Fundamentals to analyze the fundamentals of Tai Chi Chuan's movements in order to develop the methods of relaxation from Tai Chi Chuan's principle movement movements It also shows that the two techniques have commonalities in many ways. First of all, taking a philosophical approach on the body movements of Tai Chi Chuan and Bartenieff, for both methods the ultimate goal is the integration of mind and body. In other words, there is a thread of connection between the East's body and mind monism and the west's Body Awareness. Secondly, looking at it from a Breath Support standpoint as used in the Bartenieff method, the two methods both use the breathing to naturally move the body and relax the body. In Tai Chi Chuan the Breath is the basis of life and the strength of the Body. So the breathing of Tai Chi chuan is what makes body and mind communicate, harmonize and integrate. In other words, Breathing in Tai Chi is realized through mental fusion and affects the movements. This is the same as the Breath Support of Bartenieff. It is said that in every aspect the Breath Support of Bartenieff influences the movement and changes both the inner and outer form of the body. Thirdly, looking at the Core Support used in the Bartenieff method, both methods emphasize core. At the same time of moving and being conscious of one's core, the usage of muscles can be deeper rather than superficial and this enables strong and flexible movement. In Tai Chi Chuan abdominal muscles used when one coughs are consciously engaged through abdominal breathing and so strength is collected in the core. When one exercises like that the core becomes more stable and breathing becomes more smooth. Fourthly, analyzing the Rotary Factor used in the Bartenieff Fundamentals, they both use rotary movement to reach the goal of physical relaxation. The rotation factor of Bartenieff allows movement to be easier and more free because of the characteristic of joint exercise where the center axis moved in three dimensions, this is the same in Tai Chi chuan. According to Tai Chi chuan's circle and Spiral Movements, it can achieve the relaxation through switching into a seamless flow and access space as much as possible. Finally, when looking at Developmental Patterning through Bonnie Bainbridge Cohen's Body-Mind Centering Work theory, presented from Bartenieff developmental model are similar with the developmental process of Tai Chi chuan Breath, Core-Distal Connectivity/Navel Radiation, Head-Tail Connectivity/Spinal Movement, Upper-Lower Connectivity/Homologous, Body-Half Connectivity/Homo-Lateral Connectivity, Cross-Lateral Connectivity/Contra-Lateral Connectivity. They are all similar. In other words, in Tai Chi Chuan energy is gathered in the core through breathing, upper and lower body are connected through the spine, not only homo-laterally but also cross-laterally. Through this study the expression of the dance movements can be more natural. Additionally based on the Body Awareness balance usage of the central axis, joints and body can develop the relax technique.

• Annual Conference on Human and Language Technology
• /
• 1989.10a
• /
• pp.197-202
• /
• 1989

한글에서 발생하는 자소의 접촉에 의한 오인식을 해결하기 위하여, 접촉점을 중심으로 원소를 변환하여, 확장 가능한 구조를 모두 검증하고, 수락된 후보패턴에 대하여 가장 근접한 패턴 클래스로 할당하는 한글 인식방법을 제안한다. 프로그램드 배열운법을 이용하여 화소의 2차원 배열에서 입력패턴을 인식하고, PEACE(Primitive-Extraction and Attribute-Computation Embeded ) 파싱을 이용하여, 원소(primitive)의 추출과 숙성 (attribute)의 계산을 구문분석 과정에 통합하고, 전체 시스템이 동적인 구조를 갖게하여, 1차원 스트링으로의 변환에 따르는 패턴의 변형과 부가적인 노력을 억제한다.

• Journal of KIISE:Software and Applications
• /
• v.26 no.6
• /
• pp.791-803
• /
• 1999

본 논문은 두 장의 스테레요 영상으로부터 자동적으로 특징점 정합을 수행하도록 하는 새로운 절차의 효율적인 정합방법을 제안한다. 이를 위해 초기정합의 결과로 얻을 수 있는 유일 정합쌍을 이용한다. 즉, 본 논문에서는 초기정합의 결과로 얻어낸 유일 정합쌍의 정보를 이용하여 바로 outlier들을 제거시키므로써 초기정합의 결과가 갖는 애매성까지도 동반하여 상당량을 줄이도록 한다. 결국 애매성 제거에 대한 부담이 줄어들게 되므로 애매성 제거과정에서는 이완화 방법을 사용하지 않고 빠르게 애매성을 제거시킨다. 아울러 정합의 정확도를 높이기 위해 초기정합 후 바로 서브픽셀 정확도의 정합을 수행하며 정합의 마지막 단계에서는 추가정합을 수행하므로써 정합의 성능을 향상시킨다. 실내, 실회 스테레요 영상에 대한 다양한 실험결과는 본 논문에서 제안하는 방법의 특징점 정합기법이 빠르고 효율적임을 보여준다.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 1994.04b
• /
• pp.454-457
• /
• 1994

회전성형가공(Rotation molding)은 사출성형등에 비해 제품의 크기, 형상에 제약이 거의 없고 동일한 설비에서 금형의 교한만으로 다양한 제품을 생산할 수 있기 때문에 최근 국내외적으로각광을 받기 시작 하고 있다. 회전성형가공은 금형의 외부에 고온의 열을 가하면서 2축방향으로 동시에 회전을 시켜 금형 내부의 수지를 금형내부에 균일하게 용착시켜 제품을 생산하는 과정으로, 일반적으로 제품의 강도보다는 제품치수의 정도 및 치수(예를들면 두께)의 균일성이 가공상의 중요한 문제가 된다. 본 연구에서는 회전성형프로세스의 최적제어를 위한 목적으로금형의 회전운동에 관해 검토한다. 먼저 금형의 회전운동에 관한 수학적모델을 작성하고 금형 내부의 수지가 균일한 운동을 할수 있는 금형의 회전 경로를 생성하는 2축 의 회전비의 관계를 구하는 방법에 관해 검토하였다. 다음 시뮬레이션을 통하여본 방법의 타당성에 대해 검토하였다.

• Proceedings of the KIEE Conference
• /
• 2005.07a
• /
• pp.899-901
• /
• 2005

EMTP 프로그램으로 뇌써지 발생시 가공송전선로 등가모델 구현을 위한 필수적인 요소들과 비선형 요소들을 시간영역에서 해석이 가능하다. 따라서 이번 연구에서는 EMTP를 이용하여 전력전송이 이루어지는 가공 송전 선로 등가모델을 구현하고 전송선로의 길이와 굵기, 뇌격전류의 크기를 변화시켜가면서 시뮬레이션을 통해 자료를 얻는다. 이 과정으로 얻어진 결과는 송전계통의 설계에 사용되어 안정적인 전력 전송에 적용될 수 있을 것이다.

• Proceedings of the KIEE Conference
• /
• 2005.07d
• /
• pp.3111-3113
• /
• 2005

EMTP 프로그램으로 뇌써지 발생시 가공송전선로 등가모델 구현을 위한 필수적인 요소들과 비선형 요소들을 시간영역에서 해석이 가능하다. 따라서 이번 연구에서는 EMTP를 이용하여 전력전송이 이루어지는 가공 송전 선로 등가모델을 구현하고 전송선로의 길이와 굵기, 뇌격전류의 크기를 변화시켜가면서 시뮬레이션을 통해 자료를 얻는다. 이 과정으로 얻어진 결과는 송전계통의 설계에 사용되어 안정적인 전력 전송에 적용될 수 있을 것이다.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.8 no.2
• /
• pp.178-183
• /
• 1984

The stresses in lubricants by external force lead to rise in temperature and drop in viscosty, and the performance of lubricants decrease by this phenomena. The processes of shear stress generation and relaxation are linear under light load condition but those are changed into nonlinearly over a certain limit of load and speed, and this phenomena influences to viscosty change. This study investigates dense fluid which carries property change for high shear rate by using molecular dynamics, and that result can be related to research a behavior of property change of lubricants under high speed and heavy load.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.18 no.4
• /
• pp.1009-1018
• /
• 1994

An attempt is made to develop a kind of hybrid numerical method for computations of the thermal stresses during a solidification process. In this algorithm, the phase-change heat transfer analysis is perrformed by a finite volume method(FVM) and the thermal stress analysis in a solidifying body by a finite element method(FEM). The temperatures at the grid points calculated in the heat transfer analysis are transferred to those of gauss points in elements by a bi-cubic surface patch technique for the thermal stress analysis. A hyperbolic-sine constitutive law is used to prescribe the inelastic strain rate of material. Results for the unidirectional solidification process of a pure aluminum are compared with those of others and shows good agreement.