-
—組み込みプロセッサによる音響処理モジュールの開発—
松坂 要佐, 於久 健太郎, 清水 健二, 小林 哲則
p.
251
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
In this paper, we discussed about a system architecture of humanoid robot which is a large scale integrated system, to realize extensibility without real-time performance and data sharing between functional modules. We propose a multiprocessor architecture, which processes every functional module at its exclusive processor in parallel, and connects them by data shared bus. We evaluate the efficiency of the proposed architecture through the experience in implementing a multi-modal conversation robot.
抄録全体を表示
-
大庭 隆伸, 鈴木 基之, 伊藤 彰則, 牧野 正三
p.
252
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
This paper describes a dialog system for autonomous mobile robot. As the robot is battery-powered, the dialog system needs low power-consuming device. Our system uses SH-4 based speech recognizer/synthesizer board. Human-machine dialog experiments showed that the system accepted 90% of the dialogs correctly.
抄録全体を表示
-
平岡 省二, 桑野 裕康, 二矢田 勝行
p.
253
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
We introduce activities for developing and applying our speech recognition technologies to Matsushita’s products, and especially, man-machine interface using speech recognition by a remote control for a digital Hi-vision TV with AV-HDD which has many functions for services.
抄録全体を表示
-
渡辺 隆夫, 磯 健一, 磯谷 亮輔, 畑崎 香一郎, 羽金 廣, 吉田 和永
p.
254
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
We have developed a compact large-vocabulary continuous speech recognition system. As an application we have developed a Japanese/English bi-directional speech translation system for travel conversation that runs on a Personal Digital Assistant. The system translates a wide variety of spoken expressions in a travel situation. We also developed a platform for multi-modal user interface with speech and GUI interface which combines speech and data communication functions on a mobile network.
抄録全体を表示
-
石川 泰, 鈴木 忠, 河原 健太, 城戸 恵美子
p.
255
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
Speech recognition realizes efficient human-machine interface for car navigation systems with hands-free and eyes-free operations. However, since a car navigation system has limited computational power and memory, and speech recognition has also restriction, it is very difficult to design good speech interface. In this paper, we describe speech recognition for a small hardware, and show examples of multi-modal interface and its evaluation.
抄録全体を表示
-
中嶋 宏, 牛田 博英, 森 弘之, 渡辺 泰久, 齊藤 ゆみ, 石田 勉
p.
256
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
In an information-oriented society, people need to utilize various kinds of information effectively and efficiently. Today many types of information appliances are available to assist the user. But, few people can make use of the appliances to retrieve useful information. As a solution for this problem, verbal interaction systems have been developed employing information retrieval functionality and a multi-modal interface.
抄録全体を表示
-
—タンパク質構造解析自動化システムへの応用—
谷川 民生, 大場 光太郎, 平木 雅彦, 永井 稔, 若槻 壮市
p.
257
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
In this paper, we discuss about an autonomous robot system to operate a protein crystal for protein structure analysis in micro-environment. The system has a micro manipulation system and a special vision system. We mention problems in current analysis process from a point of view of high throughput. Based on the considerations, specifications for automation of the analysis process are discussed.
抄録全体を表示
-
∼IQM値を考慮に入れた可視化手法∼
大原 賢一, 水川 真, 大場 光太郎, 谷川 民生, 滝 克彦
p.
258
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
In this paper, we discuss about a vision system on autonomous robot system to operate a protein crystal for protein structure analysis in micro-environment. In the case of lucid objects, it is quite difficult to get 3D shape information from images. We propose several methods to reconstruct clear object shape using different focus distance images.
抄録全体を表示
-
新井 史人, 酒見 敏弘, 丸山 央峰, 市川 明彦, 福田 敏男
p.
259
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
In various field such as decomposition of dioxin, it’s really expected to utilize activity of microbes. But it’s difficult to monitor the microbe and to identify proper incubation conditions. So we studied on a new system to make it possible quickly and easily by using a bio-microchip. In this paper, we report about this system and the Synchronized Laser Micromanipulation(SLM) which we proposed for the trajectory control of the multiple Micro-tools by the single laser.
抄録全体を表示
-
増田 峰知
p.
260
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
Many micromanipulation devices are developed to aim at the biotechnology, medicine, and so on. However, device researchers who are almost the mechanical engineers cannot easily obtain the information on those fields. The Mie Prefectural Science and Technology Promotion Center is a local public institute that has 5 research divisions besides the industry. This report shows the investigation of the micromanipulation work seen in each division.
抄録全体を表示
-
鶴田 正也, 吉留 忠史, 河原崎 徳之, 西原 主計, 田中 孝典, 伊田 順一
p.
261
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
This paper describes a method of measuring two-dimensional position of an IC tag. This is simple technique that turns an antenna with directivity and calculates a direction and a distance from a reaction range of a radio wave sent by an IC tag. It is easy to introduce this system in our house in order not to need to install and adjust beacons on a wall. This is available for a lost article search, goal of a mobile robot, making an environment map and so on.
抄録全体を表示
-
西原 主計, 小林 繁, 河原崎 徳之, 吉留 忠史
p.
262
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
In order to direct movement of welfare vehicles etc., a position and a direction are deduced by irradiating an optical (they being infrared rays this time) beam, receiving by two or more photo-diodes, and letting a neural network pass.
抄録全体を表示
-
高 強, 西原 主計, 吉留 忠史
p.
263
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
In this paper, a voice control powered-wheelchair is developed. The powered-wheelchair has a GUI(Graphic User Interface), which is developed base on Java. In response to voice commands, sign language and text and displayed in the GUI. A voice recognition method is developed. And a field test was done at an older-aged home.
抄録全体を表示
-
濱田 彰一, 間野 隆久
p.
264
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
A robot system, which supports the work to prevent the expansion of the disaster and restore the damaged place, was developed as one of projects for reexamining national management of nuclear emergency preparedness on the occasion the critical accident in the uranium process plant occurred in 1999. The outline of the project is described below.
抄録全体を表示
-
細田 祐司, 山本 広志, 服部 誠, 岩本 太郎, 大和田 政孝, 菅野 明弘
p.
265
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
A remote control robot has been developed for disaster-prevention works in nuclear facilities. The robot can move through narrow buildings having stairs and doors by virtue of its transformable crawler mechanism, and can perform various tasks like valve operations using its remotely exchangeable work tools. Its performance was confirmed by using a mockup environment of a nuclear plant.
抄録全体を表示
-
磯崎 芳史, 中井 公一
p.
266
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
Mitsubishi Heavy Industries, Ltd. has developed two robots(MARS-A and MARS-T) for nuclear accident prevention systems. MARS-A is a work robot with a manipulator and MARS-T is a transport robot for heavy materials.
抄録全体を表示
-
大滝 真人
p.
267
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
Two types of environmental monitoring robots were developed to assist in radiation-protection measure by collecting information around the nuclear accident spots. Maps around the accident spot are made from 3-dimensional data collected by the laser range finder on the robot. Results of environmental monitoring are plotted on the maps for easy understanding of the situations. The robots have many capabilities of monitoring and are practical with functions of transportation and telecommunication.
抄録全体を表示
-
岡 潔, 三浦 弘樹, 岩井 陽典, 大日方 五郎, 竹内 浩
p.
268
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
We have developed a Radiation-proof Robot(called RaBOT) operated under the severe conditions such as gamma radiation. The robot has been operated from the operation desk with the master arm system. The master arm system is the armrest type 6 axes joystick with force feedback. However, the system has become large and complicated because the master arms have the same number of active joints as the slave manipulators; moreover, that provide not only the required sensory feedbacks on the operator but also uncomfortable constraints. This report examines the miniaturization and carriage ability for the master arm system and describes the basic performance test results.
抄録全体を表示
-
小林 忠義, 宮島 和俊, 柳原 敏
p.
269
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
The Japan Atomic Energy Research Institute (JAERI) developed the Remote Surveillance Squads (RESQs) to collect information at a nuclear facility accident site. On the basis of conceptual design studies, three remote surveillance robots (RESQ-A, B and C) were manufactured; RESQ-A is a portable vehicle to collect information for early deployment, RESQ-B is a tracked type vehicle for information collection in wide variety items, RESQ-C is a tracked type vehicle for information collection with light duty tasks. By function tests, it was confirmed that RESQs could be operated as designed.
抄録全体を表示
-
天野 久徳, 大須賀 公一, 菊池 孝司, 岩野 優樹
p.
270
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
We discuss a conceptual design of rescue robots against nuclear plant accidents. We claim that the multiple mobile small robots group is useful as rescue robots, especially in nuclear plant accidents. In this paper, we propose a new concept of rescue robot system against nuclear plant accidents. The system consists of multiple small simple robots. Moreover, we propose two robotic system robotic wall system and traction robot system.
抄録全体を表示
-
—人体姿勢マニピュレーションアルゴリズムの実装—
岩野 優樹, 大須賀 公一, 天野 久徳, 菊地 孝司
p.
271
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
This paper studies the rescue robots to rescue people in an area polluted with radioactive leakage in nuclear power institutions. In particular, we propose a rescue strategy using small traction robots and a method of manipulation algorithm of human body attitude in rescue activity, further we confirm its validity by implementation of algorithm.
抄録全体を表示
-
—視線と環境計測に基づくユーザの行動推定—
足立 佳久, 後藤 健志, 松本 吉央, 小笠原 司
p.
272
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
We are developing a control assistant system for a robotic wheelchair as a guide. Since the system can detect where the user is looking at and where the user is paying attention from the measurements of gaze and environment, it can estimate behavior of the user on the wheelchair analyzing the relations of these information.
抄録全体を表示
-
榎本 正也, 関森 大介, 宮崎 文夫
p.
273
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
A self-localization scheme for mobile robots based on model-fitting of a floor region captured by omni-directional vision is proposed. Even if parts of the floor region hidden by obstacles and walls, these can be compensated for by computing improved convex hull of a boundary point set of the floor region and by integrating multiple floor regions on the omni-directional images.
抄録全体を表示
-
根岸 善朗, 三浦 純, 白井 良明
p.
274
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
This paper describes a method of controlling the robot speed using a probabilistic occupancy map. If the robot is not sure that the target region is free, the robot has to observe it several times. We develop a method to determine an optimal speed considering the distance to the region and the uncertainty of map and observation.
抄録全体を表示
-
坪井 辰彦, 平井 慎一
p.
275
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
Detection of the position and the orientation of planar motion objects is a key to advanced object handling. We will propose a detection method using the Radon transform and phase-only correlation of Fourier transform. The proposed vision algorithm performs Radon transform of sample image and input image. Next, this algorithm compares one dimension Fourier transform of Radon transforms. Then, this algorithm detects the position and the orientation of a planar motion objects using phase-only correlation.
抄録全体を表示
-
本田 久平, 桐木 利弘, 長谷川 勉
p.
276
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
We describe a real-time vision system to measure the pose of an object grasped by a multi-fingered hand. The system consists of two sets of stereo cameras; one is mounted on the palm of the hand, and another is mounted outside the hand. This system solves the problem that the grasped object is partially visually occluded by fingers or other objects, and also improves the accuracy of pose estimation.
抄録全体を表示
-
岡本 和也, 山口 健吾, 丸 典明
p.
277
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
We propose that developed a humanoid robot. Humanoid robot carrying the active stereo camera. It uses linear visual servoing. Linear visual servoing is a system which does not depend on camera position, it is effective to the mobile robot carrying the rotated type cameras.
抄録全体を表示
-
加賀美 聡, カフナー ジェームズ, 西脇 光一, 稲葉 雅幸, 井上 博允
p.
278
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
-
宮下 敬宏, 石黒 浩
p.
279
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
In this paper, we describe an architecture and a control method for multi degrees of freedom (DOFs) of an wheeled inverted pendulum type of humanoid robot, Robovie-III. The wheeled inverted pendulum type robot can move in rough terrain with robust disturbance and save own energy by acquiring balance and maintaining stability. A dynamics of the robot is, however, changed by motion of the upper half of the body of it. Thus, in our method, we control the motion under controlling dynamic parameters of it.
抄録全体を表示
-
小林 秀行, 遠藤 修司
p.
280
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
-
田中 隆, 田中 文夫
p.
281
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
-
三井 英樹, 西村 崇
p.
282
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
Ridual is a Web site editor based on XML/Java. It describes a Web site using GUI (Graphical User Interface), analyzes existing site, and manages information in XML. Ridual functions as a reporting tool to grasp current condition; in analyzing of existing sites, it checks not only links between pages but also links and use of resource situation such as image, etc. It assists day-to-day operations. It works effectively as a reporting tool for the present situation grasp of a web site.
抄録全体を表示
-
本橋 聖一, 伊藤 靖文, 堺 浩, 近藤 浩一
p.
283
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
-
(大規模点群処理によるリバースエンジニアリング)
桜井 裕
p.
284
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
-
三ツ井 茂, 竹村 伸太郎, 坂口 嘉之
p.
285
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
-
稲邑 哲也, 中村 仁彦
p.
286
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
The discovery of Mirror Neurons in human brain shows that the motion recognition and generation are executed by bidirectional computation model, and the model leads to symbol grounding system through embodied humanoids. We have proposed mirror neuron models based on Hidden Markov Models. In this paper, we propose a new method for usual models in order to be developed through behavior observation and generation.
抄録全体を表示
-
岡田 昌史, 中村 大介, 中村 仁彦
p.
287
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
In this paper, we develop the hierarchical design method of the dynamics-based information processing system. This system has a sensor space that recognizes sensor signals and motor space that generates humanoid motions. These two spaces are connected with parameters that define the basin of attractors. Through the interaction with environment, the humanoid robot generates and transits its motion depending on the internal state.
抄録全体を表示
-
吉川 雄一郎, 浅田 稔, 細田 耕
p.
288
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
Building a robot which acquire body scheme/body image by itself is one of the most intersting issues which aims to build an intelligent robot and understand human intelligence. In the existing work, the robot designer gives the explicit representation of the robot body and needs to re-calibrate it when it has to adapt itself to the changes in its bodies or the environment. In this paper, we propose a method for cross modal map generation which learns constancy of the correlation among multi-modal sensors. The generated map could be regarded as a representation of body surface. A preliminary experiment is shown and future issues are discussed.
抄録全体を表示
-
第1報 モジュラステートネットを用いた行動制御
高橋 秀行, 仲谷 篤人, 小林 啓吾, 潮 俊光
p.
289
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
In this report, we propose a supervisory control system for motion planning of humanoid robots. The proposed system is hierarchically structured into two levels. The upper system takes charge of the motion planning, and the lower system controls and monitors the motion. In this first report, we explain the lower system. The lower level controls and monitors the robots using modular state nets.
抄録全体を表示
-
第2報 時間付きペトリネットを用いた行動計画
仲谷 篤人, 高橋 秀行, 小林 啓吾, 潮 俊光
p.
290
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
In this report, we propose a supervisory control system for motion planning of humanoid robots. The proposed system is hierarchically structured into two levels. The upper system takes charge of the motion planning, and the lower system controls and monitors the motion. In this second report, we explain the upper system. The upper level generates an optimal sequence of actions using timed Petri nets.
抄録全体を表示
-
土屋 和雄, 青井 伸也, 辻田 勝吉
p.
291
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
This paper proposes a new method of locomotion control of a biped locomotion robot. The control system composed of joint controllers and nonlinear oscillators. The oscillators, which generate reference trajectories of the joints, tune the phases with the signals from the touch sensors. As a result, a stable locomotion corresponding to the properties of the environment is established.
抄録全体を表示
-
安田 正信, 神野 雄三郎, 堤 幹夫, 百田 均
p.
292
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
一般にロータリジョイントは、回転軸に取り付けて流体と電気を回転接続します。ロボット手首用ロータリジョイントは、ロボット手首の最終回転軸に取り付けて、ロボットのベースからのホースやケーブル(配管配線)を手首部分で回転接続させてエンドエフェクタ(ハンド等)につなげます。ロボット手首用ロータリジョイントを使用する事で、主に以下のメリットがあります。·配管配線の余分な弛みを無くし、配管配線によるトラブルを解消すると共に、ロボットの能力(可動範囲)を制限させること無く、有効に利用できます。·また、ロボットが複雑な動作をしても、配管配線の動きは小さく単純になる事で、オフラインティチングや、コンピュータ上でシミュレーションできるようになります。
抄録全体を表示
-
岡本 浩幸, 西久保 賢司, 加藤 大一郎, 増田 裕康
p.
293
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
-
日台 健一, 岡部 公輔, 溝口 博, 平岡 和幸, 重原 孝臣, 三島 健稔, 吉澤 修治
p.
294
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
This paper presents an working implementation of face recognition system that is combination of robust face detector and interactively trainable face identifier. The face detector is robust against size variation and brightness fluctuation. The identifier is based on on-line linear discriminant analysis(OLDA), which is an extension of conventional LDA. Experimental result supports feasibility and effectiveness of the implemented system.
抄録全体を表示
-
岡部 公輔, 日台 健一, 平本 真琴, 溝口 博, 平岡 和幸, 重原 孝臣, 三島 健稔, 吉澤 修治
p.
295
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
This paper presents an implementation of pattern recognition system that is possible to discriminate arbitrary patterns. The advantage of the implemented system is that we can easily add a new class of pattern through GUI while the system is running. This interactive trainability is realized by on-line linear discriminant analysis (OLDA). It reduces heavy calculation that conventional LDA requires.
抄録全体を表示
-
篠原 徹也, 岡田 昌史, 中村 仁彦
p.
296
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
We developed a new hip joint mechanism. Double spherical hip joint consists of two spherical joints whose center of rotations agree at one point and can realize two functions. One function is as the hip joint and the other function is as the waist joint. It is expected that we can controll the balance of the humanoid easily by taking advantage of this mechanism.
抄録全体を表示
-
Lekskulchai PONGSAK, Masafumi OKADA, Yoshihiko NAKAMURA
p.
297
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
In controlling a humanoid robot, state estimators are used to estimate condition of motion and orientation of the body part which will be feedbacked to control systems to regulate level of the body along with controlling trajectory of center of gravity (C.G.) of the robot. In this research, an algorithm for state estimators is developed by applying the H2-norm optimal filtering technique to a linearized dynamic model of the body part of the robot.
抄録全体を表示
-
G. Chevant, Y. Nakamura
p.
298
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
The development of a kinematical/dynamical model of a humanoid robot to be used for simulation is considered. The aim of this paper is to formalize the process to produce the model containing all the required data in order to simulate the behaviour of a humanoid in its environment. Then the way to organise the data and to choose the references depends on the simulator used. An application for OpenHRP with the humanoid robot of the NOL Laboratory is proposed here.
抄録全体を表示
-
鈴木 一郎, 栗原 一貴, 中村 仁彦
p.
299
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
This paper describes a way to develop a high precision motion capture system for detailed human model. We discuss each process of optical motion capture sequence with many reflective markers and detailed human model. And we propose a method to calibrate kinematic parameters of human model and placement of markers attached on body.
抄録全体を表示
-
中島 秀之, 橋田 浩一
p.
300
発行日: 2002年
公開日: 2003/05/15
会議録・要旨集
フリー
The goal of the project is to develop human-centered information processing assistance systems (intelligence booster) usable without special knowledge or training, while addressing the problems of information overload and privacy. We are aiming to construct a system that can provide situated information that I-want-here-now through “natural interface.” In other words, we are strengthening a variety of technologies that link digital realm represented by Internet to us people who live in the real world.
抄録全体を表示