음성-영상 융합 음원 방향 추정 및 사람 찾기 기술

최종석(JongSuk Choi),이병기(Byoung-gi Lee),윤상석(SangSuk Yoon),최문택(Mun-Taek Choi),김문상(Munsang Kim),김대진(Daijin Kim) 대한기계학회 2010 대한기계학회 춘추학술대회 Vol.2010 No.11

Service robot is equipped with various sensors such as vision camera, sonar sensor, laser scanner, microphones, and the like. Although these sensors have their own functions, some of them can work together to perform more complicated functions. Audio-vision fusion is a typical and powerful combination of two sensors because audio is complementary to vision and vice versa. Human-beings also mainly depend on visual and auditory information in their daily life. In this paper, we provide two studies using audio-vision fusion. One is about enhancing performance of sound localization. The other is about robot attention with sound localization and face detection.

외부 교란하에 있는 이족 보행 로봇의 균형 회복 전략

이상용(Sang-yong Lee),김문상(Mun-sang Kim),최문택(Mun-taek Choi) 대한기계학회 2010 대한기계학회 춘추학술대회 Vol.2010 No.11

This paper first starts from the concept that momentum generated from COM of robot is an important parameter of keeping the balance of robot, and the robot will not fall down as long as the COM momentum is maintained as zero. However, it is difficult to measure this momentum with sensor in the actual robot where external perturbation is working, and even more difficult to make it into zero. In this study, LIDM oriented from LIPM is suggested. By this model, angular momentum generated in COM actually can be predictive. Additionally, we define that ZMP regions are divided into 3 stages by the strength of external perturbation in order to avoid robot’s fall. Hence, balance recovery region defined in each stage signifies the minimum radius of action for avoiding robot’s fall. This becomes an important criterion of proposing a solution to act upon each stage differently so that the robot may not actually fall down by external perturbation.

인간의 비언어적 행동 특징을 이용한 다중 사용자의 상호작용 의도 분석

윤상석(Sang-Seok Yun),김문상(Munsang Kim),최문택(Mun-Taek Choi),송재복(Jae-Bok Song) 제어로봇시스템학회 2013 제어·로봇·시스템학회 논문지 Vol.19 No.8

According to the cognitive science research, the interaction intent of humans can be estimated through an analysis of the representing behaviors. This paper proposes a novel methodology for reliable intention analysis of humans by applying this approach. To identify the intention, 8 behavioral features are extracted from the 4 characteristics in human-human interaction and we outline a set of core components for nonverbal behavior of humans. These nonverbal behaviors are associated with various recognition modules including multimodal sensors which have each modality with localizing sound source of the speaker in the audition part, recognizing frontal face and facial expression in the vision part, and estimating human trajectories, body pose and leaning, and hand gesture in the spatial part. As a post-processing step, temporal confidential reasoning is utilized to improve the recognition performance and integrated human model is utilized to quantitatively classify the intention from multi-dimensional cues by applying the weight factor. Thus, interactive robots can make informed engagement decision to effectively interact with multiple persons. Experimental results show that the proposed scheme works successfully between human users and a robot in human-robot interaction.

가사활동을 돕는 서비스 로봇(Personal Robot)의 시스템 설계

주정현(Junghyun Chu),노학종(Hakjong Noh),최문택(Mun-Taek Choi),김문상(Munsang Kim),고인영(Inyoung Ko),박종우(Chongwoo Park) 대한기계학회 2010 대한기계학회 춘추학술대회 Vol.2010 No.11

This paper aims one system to combine components that is Personal Robot to help housekeeping in complex environment of house. The needed skills are Cognitive environment, Navigation, Mobile Manipulation, Object recognition and FT(Force Torque) Sensor Value Control. This research is used by 2-differential-wheels platform robot. Various components make in one system that is difficult, because to make the perfect function is not easy. So this thesis describes how make up for a few problems is come from integrating of components.

음성-영상 융합 음원 방향 추정 및 사람 찾기 기술

이병기(Byoung-gi Lee),최종석(JongSuk Choi),윤상석(SangSuk Yoon),최문택(Mun-Taek Choi),김문상(Munsang Kim),김대진(Daijin Kim) 대한기계학회 2011 大韓機械學會論文集A Vol.35 No.7

서비스 로봇은 비전 카메라, 초음파 센서, 레이저 스캐너, 마이크로폰 등과 같은 다양한 센서를 장착하고 있다. 이들 센서들은 이들 각각의 고유한 기능을 가지고 있기도 하지만, 몇몇을 조합하여 사용함으로써 더욱 복잡한 기능을 수행할 수 있다. 음성영상 융합은 서로가 서로를 상호보완 해주는 대표적이면서도 강력한 조합이다. 사람의 경우에 있어서도, 일상생활에 있어 주로 시각과 청각 정보에 의존한다. 본 발표에서는, 음성영상 융합에 관한 두 가지 연구를 소개한다. 하나는 음원 방향 검지 성능의 향상에 관한 것이고, 나머지 하나는 음원 방향 검지와 얼굴 검출을 이용한 로봇 어텐션에 관한 것이다. Service robots are equipped with various sensors such as vision camera, sonar sensor, laser scanner, and microphones. Although these sensors have their own functions, some of them can be made to work together and perform more complicated functions. AudioFvisual fusion is a typical and powerful combination of audio and video sensors, because audio information is complementary to visual information and vice versa. Human beings also mainly depend on visual and auditory information in their daily life. In this paper, we conduct two studies using audioFvision fusion: one is on enhancing the performance of sound localization, and the other is on improving robot attention through sound localization and face detection.

건설-제조 현장 작업자의 복합 질환 선별을 위한 건강 검진 데이터를 사용한 이진-관련성-기반 다중 레이블 분류

등덕호(Teh-Hao Teng),이태림(Tae-Rim Lee),변승혁(Seung Hyeok Byeon),최문택(Mun-Taek Choi) 한국정보기술학회 2024 한국정보기술학회논문지 Vol.22 No.2

The objective of this study is to explore the potential of building an artificial intelligent model with medical examination data from construction site workers to predict health status in terms of multiple illnesses. The data set, collected from 601 field workers, consists of 82 features and 9 illness labels. Given that an individual may be diagnosed with multiple illnesses, the AI model adopts a binary relevance-based multi-label classification approach. Various classification algorithms were trained through grid search-based hyperparameter tuning to select the optimal model. This is the first study to apply artificial intelligence models using large-scale medical examination data to identify suspected complex illnesses in construction and manufacturing workers.

로봇을 이용한 장애아동 원격지원 시스템 구축

김창걸,이효신,신윤희,강정배,김문상,최문택,송병섭 대구대학교 특수교육재활과학연구소 2012 특수교육재활과학연구 Vol.51 No.1

교육과 치료를 위해 많을 시간을 투자하여야 하는 장애인들이 시간과 공간적 제약에서 벗어나 각종 서비스를 원하는 곳에서 원하는 시간에 받을 수 있다면 그들의 삶의 질은 향상될 것이다. 따라서 본 연구에서는 국내 로봇기술과 통신 기술을 조합하여 장애아동에게 원격지원 중재가 가능한 로봇 원격지원 시스템을 구축하였으며, 구축된 시스템을 D시 장애 전담 어린이집 2곳과 K시 특수교육대학에 1곳에 설치하였다. 또한 구축된 시스템을 통한 중재 가능성을 알아보기 위하여 구축된 시스템을 사용하여 장애아동에게 원격지원 중재를 적용한 특수교육 전문가 8명을 대상으로 시스템의 기능별 만족도 조사 하였으며, 학령기 아동 6명 대상으로 구축된 시스템을 사용하여 대화를 통한 상호작용이 가능한지 알아보았다. 그 결과 전문가들은 대부분의 로봇원격지원 시스템의 기능에 대하여 5점 만점에 평점 3.5 이상으로 만족 하였으며, 대화를 통한 상호작용 또한 일반교사와 비교하여 차이가 나지 않음을 알 수 있었다. 하지만 시스템의 반응속도와 팔의 움직임, 안전에 관한 부분에서는 불만족한 의견을 나타내었다. 향후 이러한 문제점을 해결하기 여러 곳에 사용가능한 범용 시스템이 아니라 중재 분야별 필요기능만 탑재한 맞춤형 시스템을 제작하여야 할 것이다. The disabled must spend a lot of time on their education and treatment. If it is possible for them to break free from constraints of time and space and to receive each service no matter where or when, their quality of life will be improved. Thus, this study established a robot remote support system that can provide children with disabilities with remote support therapy by combining domestic robot technology and communications technology, and the established system was installed at the two exclusive nursery schools for disabled children in D City and one department of special education in K City. In addition, this study conducted a satisfaction survey with each function of the system targeting 8 special education professionals who applied the remote support intervention to the children with disabilities in order to confirm the intervention possibility by the established system, and experimented whether it was possible for the six school-age children to interact by using the established system. As a result, the professionals were satisfied with most functions of the robot remote support system giving more than 3.5 out of 5.0. It also showed that the interaction by conversation was not different from that of general teachers. However, they expressed a dissatisfactory opinion on the reaction velocity of the system, movement of the arm and its safety. Henceforth, these problems should be solved, and it will be necessary to create not a general-purpose system available for various uses but a customized system equipped with only necessary functions by intervention field.