T. Kuniyoshi, M. Inaba, and H. Inoue, "Teaching by showing: Generating robot programs by visual observation of human performance," in Proc. 20th Int. Symp. Industrial Robots, 1989, pp. 119--126.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....facilitate the incorporation of sensor strategies in their XPROBE system. XPROBE would then generate AML code for the performance of the desired task. 1.1. 4 Automatic Programming The problems associated with teaching and textual programming can be alleviated by automatic programming (e.g. 55] [78], 91] 149] also known as task level programming [90] In automatic programming, conceptually, the only inputs to the robot system required to generate the control command sequences are the description of the objects involved in the task, and the high level task specifications. Realization of ....

....grasping motions. This method of task programming would obviate the need for a programmer to explicitly describe the required task, since the system is able to understand the task based on observation of the task performance by a human. An approach similar to APO was taken by Kuniyoshi, et al. [78], who developed a system which is capable of emulating the performance of a human operator using a real time stereo vision system. However, their system is restricted to pick and place operations and fingertip grasps. In order to replicate observed tasks, our system first infers intended subgoals ....

[Article contains additional citation context not shown here]

T. Kuniyoshi, M. Inaba, and H. Inoue, "Teaching by showing: generating robot programs by visual observation of human performance," Proc. 20th Int'l Symp. on Industrial Robots, 1989, pp. 119-126.

....lacks of any means to communicate this path to the robot. A CAD oriented trajectory design editor was proposed to support the explicit programming of robot paths [2] but such an editor again depends on the geometrical model of the workcell. In this paper we employ a teaching by showing paradigm [3] as an efficient approach to explicit programming of robot paths. Methods for the specification of robot poses and paths which utilize teaching by showing were proposed in [4, 5, 6, 7] In [5] vision was employed to teach the poses on the desired path at some important passing points whereas in ....

Y. Kuniyoshi, M. Inaba, and H. Inoue. Teaching by Showing: Generating Robot Programs by Visual Observation of Human Performance. In Proc. 20th Int. Symp. Industrial Robots, pages 119--126, Tokyo, Japan, 1989.

....between various types of grasps. 1 Chapter 1 Introduction Robot programming is an essential component of task automation. The current methods for robot programming include teaching (e.g. 1] 2] textual programming (e.g. 3] 5] and automatic programming (e.g. 4] 6] 8] [11]) The first two methods are by far the most pervasive in both the industrial and academic environments. In teaching methods, the robot or manipulator learns its trajectory either through a teach pendant or actual guidance through the sequence of operations ( teach by guiding or less ....

....the task from observation, and generates the same assembly sequence for a robot. In this paradigm, the human operator does all the thinking the system understands what needs to be done based on what is observed and performs the task or tasks. A similar approach was taken by Kuniyoshi et al. [11] who developed a system which emulates the performance of a human operator. However, their system is restricted to pick and place operations. This report is an account of our work on grasp identification. In Chapter 2, we describe a 3 D structure called the contact web which can be used to ....

Kuniyoshi, T., Inaba, M., and Inoue, H., "Teaching by Showing: Generating Robot Programs by Visual Observation of Human Performance," Proc. 20th Int'l Symp. on Industrial Robots, 1989, pp. 119-126.

....required for a simple pick and place example Visual programming does little for the bulk of robotic applications that involve substantive contact with the environment, uncertainty, and complex sensing. A more recent approach to human robot interaction is the field of learning by observation [7] [8]. By forcing the robot to observe a human interacting with the world, rather than forcing the human to interact with a representation of the robot interacting with the world, a more natural, rich, anthropocentric environment results. But these systems are mostly kinematically based and operate ....

Kuniyoshi, T., M. Inaba, and H. Inoue, "Teaching by Showing: Generating Robot Programs by Visual Observation of Human Performance," in Proc. of the 20th International Symp. on Industrial Robots, 1989, pp. 119-126.

....this elementary operative intelligence, which is used by humans unconsciously, in a computer controlled robot [20] This fact gives rise to an idea to directly transfer the user s implicit knowledge about the desired trajectory to the robot. In this paper we utilize a teaching by showing paradigm [18] as an efficient approach to achieve this goal. Robot motions can be classified as fine and gross motions [16] Fine motions combine motion commands and on line sensing. They are used when the robot must maintain a certain relationship with other objects, e.g. when inserting a peg into a hole. ....

....(iii) using the measurements provided by the vision system, the motion is reconstructed and transformed into an executable program. We refer to this method as teaching by showing. Vision based robot programming has received some attention in the past. Ikeuchi et al. 17] and Kuniyoshi et al. [18, 19] used this paradigm for the task specification. Both systems are based on a visual recognition of human action sequences. This enables the generation of a symbolic description of the task that can be converted into an executable robot program. However, both systems can analyze only extremely ....

Y. Kuniyoshi, M. Inaba, and H. Inoue. Teaching by showing: Generating robot programs by visual observation of human performance. In Proc. 20th Int. Symp. Industrial Robots, pages 119--126, Tokyo, Japan, 1989.

....case, the designer would have to draw the automaton model that best suits the task(s) under observation, depending on the application requirements. The code for the state machine then needs to be implemented. Automatic construction of the state machine could be done by having a learning stage [11,12], in which a mapping module would form the automaton. The building phase is performed before the actual observation process is invoked. The idea is to supply the module with sets of possible sequences in the form of strings of a certain language that the DEDS automaton should minimally accept. The ....

Y. Kuniyoshi, M. Inaba, and H. Inoue, "Teaching by showing : Generating robot programs by visual observation of human performance", 20 th ISIR, 1989.

....teaching [15] during which the robot is intuitively and expertly guided through the proper motions of the task by the expert user. Gesture Based Programming is an extension of the successful aspects of lead through teaching and also of the emerging learning by observation paradigms [6][9]. It extends these ideas by combining multi agent [1] and skillbased [11] concepts to form encapsulated expertise which guides the recognition and segmentation of gestures during teaching and guides error recovery and exception handling of the robot during execution. In a nutshell, gesture based ....

Kuniyoshi, T., M. Inaba, and H. Inoue, "Teaching by Showing: Generating Robot Programs by Visual Observation of Human Performance," in Proc. of the 20th International Symp. on Industrial Robots, 1989, pp. 119-126.

....programming environment. Visual programming does little for the bulk of robotic applications that involve substantive contact with the environment, uncertainty, and complex sensing. A more recent approach to human robot interaction is the field of learning by observation (Kang and Ikeuchi, 1996)(Kuniyoshi et al., 1989). By forcing the robot to observe a human interacting with the world, rather than forcing the human to interact with a textual representation of the robot interacting with the world, a more natural, anthropocentric programming environment results. But these systems are mostly kinematically ....

Kuniyoshi, T., M. Inaba, and H. Inoue, (1989). "Teaching by Showing: Generating Robot Programs by Visual Observation of Human Performance," Proc. of 20th Intnl. Symp. on Ind. Robots, pp. 119-126.

....by the programming of a robot trajectory by use of teach in or play back methods. Vision based programming paradigm can be employed to simultaneously solve both problems. A straightforward method for transferring the user s knowledge about the task by means of vision is Teaching by Showing [9] which can be viewed as a part of vision based programming. When using this method, the user shows the intended task to the robot by actually doing it himself. The performed task is observed and processed by the robot s visual system. The Ales Ude is on leave from the Department of Automation, ....

Y. Kuniyoshi, M. Inaba, and H. Inoue. Teaching by showing: Generating robot programs by visual observation of human performance. In Proc. 20th Int. Symp. Industrial Robots, pages 119--126, Tokyo, Japan, 1989.

....the workspace and the structure of the manipulator. In contrast, task plans described in terms of more abstract and qualitative representation of assembly operations such as pick and place are of high reusability. We have proposed a novel framework of task instruction called teaching by showing [11]. It consists of five phases: Presentation: The instructor shows the target task to the system by performing it with his hands. The instructor needs to perform the task only once from start to end without pausing. Recognition: The system observes the performance by vision and constructs a ....

Y. Kuniyoshi, M. Inaba, and H. Inoue, "Teaching by showing: Generating robot programs by visual observation of human performance", in Proc. Int. Symp. Industrial Robots, 1989, pp. 119--126.

No context found.

T. Kuniyoshi, M. Inaba, and H. Inoue, "Teaching by showing: Generating robot programs by visual observation of human performance," in Proc. 20th Int. Symp. Industrial Robots, 1989, pp. 119--126.

No context found.

Y. Kuniyoshi, M. Inaba and H. Inoue, Teaching by Showing : Generating Robot Programs by Visual Observation of Human Performance, Department of Mechanical Engineering, The University of Tokyo, Technical Report, 1990.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC