• MALSORI
• /
• no.57
• /
• pp.85-97
• /
• 2006

This paper suggests an algorithm that can estimate the direction of the sound source in real time. The algorithm uses the time difference and sound intensity information among the recorded sound source by four microphones. Also, to deal with noise of robot itself, the Kalman filter is implemented. The proposed method can take shorter execution time than that of an existing algorithm to fit the real-time service robot. Also, using the Kalman filter, signal ratio relative to background noise, SNR, is approximately improved to 8 dB. And the estimation result of azimuth shows relatively small error within the range of ${\pm}7$ degree.

• 한국HCI학회:학술대회논문집
• /
• 2007.02a
• /
• pp.688-693
• /
• 2007

본 논문은 인간과 로봇간의 효과적인 상호작용을 위한 로봇제스쳐의 자동생성 기법을 다룬다. 이는 텍스트 정보 만의 입력으로 의미 있는 단어에 대응되는 특정 제스쳐패턴이 자동적으로 생성되도록 하는 기법으로서 이를 위한 사전조사로 제스쳐가 출현하는 발화시점에서의 단어수집이 우선적으로 요구되었다. 본 논문은 이러한 분석을 위해 두 개 이상의 연속된 제스쳐 패턴을 효과적으로 표현할 수 있는 제스쳐 모델을 제안한다. 또한 제안된 모델이 적용되어 구축된 제스쳐DB와 스크립트 기법을 이용한 로봇제스쳐 자동생성 방법을 제안한다. 제스쳐 생성시스템은 규칙기반의 제스쳐 선택부와 스크립트 기반의 동작 계획부로 구성되고, 집사로봇의 안내기능에 대한 모의실험을 통해 그 효용성을 확인한다.

• Korean Journal of Computational Design and Engineering
• /
• v.12 no.1
• /
• pp.50-57
• /
• 2007

In recent years, many robotics researches have been focused on developing human-friendly robots, that is, humanoid biped robots. The researches of humanoid robots include various areas such as hardware development, control of biped locomotion, artificial intelligence, human interaction, etc. The present work concerns the hardware development of a mid-size humanoid robot, BONOBO, focusing on rapid development of outer body parts with integrated application if CAD/CAM/CAE/RP. Most parts are three-dimensionally designed using 3D CAD, and effectively connected with CAE analyses using both kinematic simulation and structural analysis. In order to reduce lead time and investment cost for parts developments, Rapid Prototyping (RP) and CAM are selectively utilized for manufacturing body parts. These master parts are then replicated using the vacuum casting process, from which we can obtain plastic parts repeatedly. Through this integrated approach, the first prototype of BONOBO can be successfully developed with relatively low time and investment costs.

• Journal of Institute of Control, Robotics and Systems
• /
• v.16 no.5
• /
• pp.415-419
• /
• 2010

A variety of methods for sound source localization have been developed and applied to several applications such as noise detection system, surveillance system, teleconference system, robot auditory system and so on. In the previous work, we proposed the sound source localization using the spatially mapped GCC functions based on TDOA for robot auditory system. Performance of the proposed one for the noise effect and estimation resolution was verified with the real environmental experiment under the single source assumption. However, since multi-talker case is general in human-robot interaction, multiple source localization approaches are necessary. In this paper, the proposed localization method under the single source assumption is modified to be suitable for multiple source localization. When there are two sources which are correlated, the spatially mapped GCC function for localization has three peaks at the real source locations and imaginary source location. However if two sources are uncorrelated, that has only two peaks at the real source positions. Using these characteristics, we modify the proposed localization method for the multiple source cases. Experiments with human speeches in the real environment are carried out to evaluate the performance of the proposed method for multiple source localization. In the experiments, mean value of estimation error is about $1.4^{\circ}$ and percentage of multiple source localization is about 62% on average.

• Journal of Semiconductor Engineering
• /
• v.2 no.2
• /
• pp.130-135
• /
• 2021

Recently, deep neural networks (DNNs) are actively used for action control so that an autonomous system, such as the robot, can perform human-like behaviors and operations. Unlike recognition tasks, the real-time operation is essential in action control, and it is too slow to use remote learning on a server communicating through a network. New learning techniques, such as reinforcement learning (RL), are needed to determine and select the correct robot behavior locally. In this paper, we propose an energy-efficient DNN processor with a LUT-based processing engine and near-zero skipper. A CNN-based facial emotion recognition and an RNN-based emotional dialogue generation model is integrated for natural HRI system and tested with the proposed processor. It supports 1b to 16b variable weight bit precision with and 57.6% and 28.5% lower energy consumption than conventional MAC arithmetic units for 1b and 16b weight precision. Also, the near-zero skipper reduces 36% of MAC operation and consumes 28% lower energy consumption for facial emotion recognition tasks. Implemented in 65nm CMOS process, the proposed processor occupies 1784×1784 um2 areas and dissipates 0.28 mW and 34.4 mW at 1fps and 30fps facial emotion recognition tasks.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2017.07a
• /
• pp.13-16
• /
• 2017

제스처 인식은 HCI(Human-Computer Interaction) 및 HRI(Human-Robot Interaction) 분야에서 활발히 연구되고 있는 기술이며, 제스처 데이터의 특징을 추출해내고 그에 따른 분류를 통하여 사용자의 제스처를 정확히 판별하는 것이 중요한 과제로 자리 잡았다. 본 논문에서는 EMG(Electromyography) 센서로 측정한 사용자의 손 제스처 데이터를 분석하는 방안에 대하여 서술한다. 수집된 데이터의 노이즈를 제거하고 데이터의 특징을 극대화시키기 위하여 연속적인 데이터로 변환하는 전처리 과정을 거쳐 이를 머신 러닝 알고리즘을 사용하여 분류하였다. 이 때, 기존의 raw 데이터와 전처리 과정을 거친 데이터의 성능을 decision-tree 알고리즘을 통하여 비교하였다.

• Proceedings of the KIEE Conference
• /
• 2007.10a
• /
• pp.117-118
• /
• 2007

A robot needs a human detection algorithm for interaction with a human. This paper proposes a method that finds people using a SVM (support vector machine) classifier and a stereo camera. Feature vectors of SVM are extracted by HoG (histogram of gradient) within images. After training extracted vectors from the clustered images, the SVM algorithm creates a classifier for human detection. Each candidate for a human in the image is generated by clustering of depth information from a stereo camera and the candidate is evaluated by the classifier. When compared with the existing method of creating candidates for a human, clustering reduces computational time. The experimental results demonstrate that the proposed approach can be executed in real time.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.8 no.7
• /
• pp.1480-1487
• /
• 2004

This study is proposed the novel gesture recognition method for the remote camera robot control. To recognize the dynamics gesture, the preprocessing step is the image segmentation. The conventional methods for the effectively object segmentation has need a lot of the cole. information about the object(hand) image. And these methods in the recognition step have need a lot of the features with the each object. To improve the problems of the conventional methods, this study proposed the novel method to recognize the dynamic hand gesture such as the MMS(Max-Min Search) method to segment the object image, MSM(Mean Space Mapping) method and COG(Conte. Of Gravity) method to extract the features of image, and the structure of recognition MLPNN(Multi Layer Perceptron Neural Network) to recognize the dynamic gestures. In the results of experiment, the recognition rate of the proposed method appeared more than 90[%], and this result is shown that is available by HCI(Human Computer Interface) device for .emote robot control.

• Journal of Digital Contents Society
• /
• v.18 no.8
• /
• pp.1501-1507
• /
• 2017

Telepresence in robot assisted learning has preferred low-height, shorter than life-size robotic platforms for reasons such as operational stability, user convenience and psychological comfort in human robot interaction. If, however, the reason for using a telepresence robot is to display the authority of a social superior to a social inferior, one can hypothesize that a robotic platform which reflects real-life height advantage would be better suited for the stated purpose than conventional low-height platforms. In order to test the hypothesis, we examined whether the height of the robot had an effect on an instructor connected to a telepresence robot in robot-assisted learning with regard to controlling a large number of elementary school students. The pre-and post experiment demonstrates that the use of a life-size telepresence robot, compared to a child-size telepresence robot, failed to make a meaningful difference in the instructors' authority being accepted by the students. However, behavioral measures shows that a taller robot has more merits in controlling students.

• 제어로봇시스템학회:학술대회논문집
• /
• 2005.06a
• /
• pp.457-462
• /
• 2005

Industrial and economical importance of CP(Cellular Phone) is growing rapidly. Combined with IT technology, CP is currently one of the most attractive technologies for all. However, unless we find a breakthrough to the technology, its growth may slow down soon. RT(Robot Technology) is considered one of the most promising next generation technology. Unlike the industrial robot of the past, today's robots require advanced technologies, such as soft computing, human-friendly interface, interaction technique, speech recognition, object recognition, and many others. In this study, we present a new technological concept named RCP(Robotic Cellular Phone), which combines RT & CP, in the vision of opening a new direction to the advance of CP, IT, and RT all together. RCP consists of 3 sub-modules. They are $RCP^{Mobility}$, $RCP^{Interaction}$, and $RCP^{Interaction}$. $RCP^{Mobility}$ is the main focus of this paper. It is an autonomous navigation system that combines RT mobility with CP. Through $RCP^{Mobility}$, we should be able to provide CP with robotic functionalities such as auto-charging and real-world robotic entertainments. Eventually, CP may become a robotic pet to the human being. $RCP^{Mobility}$ consists of various controllers. Two of the main controllers are trajectory controller and self-localization controller. While Trajectory Controller is responsible for the wheel-based navigation of RCP, Self-Localization Controller provides localization information of the moving RCP. With the coordinate information acquired from RFID-based self-localization controller, Trajectory Controller refines RCP's movement to achieve better RCP navigations. In this paper, a prototype system we developed for $RCP^{Mobility}$ is presented. We describe overall structure of the system and provide experimental results of the RCP navigation.