Becker M, Kefalea E, Mael E et al. 1999. `GripSee: A Gesture-controlled Robot for Object Perception and Manipulation'. Autom Robots, 6:203--221.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....describe experiments to evaluate the quality of a grip using the sensor measurements and a utility that allows to guide the robot arm on a desired trajectory with negligible force. 1 Introduction The robot component we describe in this paper came into existence during work on our humanoid robot [1, 2]. Dextrous grasping with a humanoid has the problem that the cameras are located at a considerable distance from where the actual manipulation takes place. Consequently, threedimensional visual analysis has significant inaccuracy. It can be reduced if cameras are mounted on the manipulator itself, ....

....courses. It can be seen that the speed signal is hardly influenced by sliding, which in turn can be clearly detected in the vibration signal. 3. 2 The robot Place figure 6 about here The following experiments have been carried out on our humanoid robot platform, which is described in detail in [1]. As depicted in figure Fig. 6 it consists of the following components: One modular robot arm with seven degrees of freedom (DoF) kinematics similar to a human arm, and a parallel jaw gripper; a dual stereo camera head with three DoF (pan, tilt, and vergence) and a stereo basis of 30 cm ....

Mark Becker, Efthimia Kefalea, Eric Mael, Christoph von der Malsburg, Mike Pagel, Jochen Triesch, Jan C. Vorbruggen, Rolf P. Wurtz, and Stefan Zadel. GripSee: A gesture-controlled robot for object perception and manipulation. Autonomous Robots, 6(2):203--221, 1999. 7

....movements (Sect. 2) We then describe in detail the computer vision techniques used for observing the grasping movement (Sect. 3) and present first experiments (Sect. 4) Section 5 gives a discussion and an outlook. 2 A Way of Imitating Grasping Movements The robot at the Bochum laboratory [2] has a kinematically redundant arm with seven degrees of freedom, which allows it to grasp an object from various direc tions (Fig. 1) A stereo camera head with three degrees of freedom is mounted on top of the arm allowing for pan, tilt, and vergence motion. The cameras yield Control Module I ....

....which stores the grasping position and direction relative to an objectcentered coordinate system. This will allow the robot to generalize the learned grip to new situations, where the object is in a different place. The required object recognition technology is readily available at our laboratory [2]. In the following, we will focus on the techniques for observing the human teacher s grasping movements. 3 Observation of Grasping Movements 3.1 Hand Tracking The hand tracking system has been described elsewhere [12] It employs motion detection, skin color analysis and a stereo cue and ....

M. Becker, E. Kefalea, E. MI, C. v.d. Malsburg, M. Pagel, J. Triesch, J. C. Vorbrfiggen, R. P. Wfirtz, and S. Zadel. GripSee: A gesture-controlled robot for object perception and manipulation. Autonomous Robots, 6(2):203-221, 1999.

....The idea is that only if the match of the best graph is sufficiently better than those of the other graphs will the match be accepted. Details of such an approach are given in Ref. 9] We have demonstrated the applicability of our approach by creating a gesture interface for an autonomous robot [20]. The robot operates in a pick and place scenario. The user points to objects with different hand postures to indicate which object to pick up and how to pick it up, or where to Fig. 6. Posture number three performed by 10 different subjects in front of ten of the used complex backgrounds. put it ....

M. Becker, E. Kefalea, E. Mae l, C. von der Malsburg, M. Pagel, J. Triesch, J.C. Vorbru ggen, R.P. Wu rtz, S. Zadel, GripSee: a gesturecontrolled robot for object perception and manipulation, Autonomous Robots 6 (2) (1999) 203 -- 221.

....visual cortex. Thus the signals that reach the visual system have already undergone a complex and poorly understood selection process. In the following we describe two experiments to study the effects of aprioribiases on unsupervised learning of visual representations using an autonomous robot [Becker et al. 1999]. Experiment 1 establishes that biasing an agent s learning towards interesting image regions can dramatically alter the character of the representations formed to reflect the aprioridefined perceptual needs of the agent. If the robot selectively learns on image patches showing motion and skin ....

Becker, M., Kefalea, E., Mael, E., v.d. Malsburg, C., Pagel, M., Triesch, J., Vorbruggen, J. C., Wurtz, R. P., and Zadel, S. (1999). GripSee: A gesture-controlled robot for object perception and manipulation. Autonomous Robots, 6(2):203--221.

....in a complex dynamic environment and demonstrating the dexterity of our robot. This application requires of course a number of other skills needed by the robot, e.g. recognition of shape and orientation of the object pointed to, grip planning and grip execution, which are discussed elsewhere [1, 2]. 3 Tracking of Head and Hands A prerequisite for the successful recognition of gestures is tracking the head and hands of the gesturing person. We combine motion, color and stereo cues to reach the robustness demanded by real world applications. Let us first consider a hand pointing to some ....

M. Becker, E. Kefalea, E. Mael, C. v.d. Malsburg, M. Pagel, J. Triesch, J. C. Vorbruggen, R. P. Wurtz, and S. Zadel. GripSee: A gesture-controlled robot for object perception and manipulation. (submitted to) Autonomous Robots: Special Issue on Perception-Based Intelligent Robots, 1998.

....movements (Sect. 2) We then describe in detail the computer vision techniques used for observing the grasping movement (Sect. 3) and present first experiments (Sect. 4) Section 5 gives a discussion and an outlook. 2 A Way of Imitating Grasping Movements The robot at the Bochum laboratory [2] has a kinematically redundant arm with seven degrees of freedom, which allows it to grasp an object from various directions (Fig. 1) A stereo camera head with three degrees of freedom is mounted on top of the arm allowing for pan, tilt, and vergence motion. The cameras yield Control Module Hand ....

....which stores the grasping position and direction relative to an objectcentered coordinate system. This will allow the robot to generalize the learned grip to new situations, where the object is in a different place. The required object recognition technology is readily available at our laboratory [2]. In the following, we will focus on the techniques for observing the human teacher s grasping movements. 3 Observation of Grasping Movements 3.1 Hand Tracking The hand tracking system has been described elsewhere [12] It employs motion detection, skin color analysis and a stereo cue and ....

M. Becker, E. Kefalea, E. Mael, C. v.d. Malsburg, M. Pagel, J. Triesch, J. C. Vorbruggen, R. P. Wurtz, and S. Zadel. GripSee: A gesture-controlled robot for object perception and manipulation. Autonomous Robots, 6(2):203--221, 1999.

....Proceedings of KES 99 at Adelaide, Australia. 1999, pp. 423 426. c #IEEE. Abstract As a step towards systems that can acquire knowledge automatically we have designed a system that can learn new objects with a minimum of user interaction and implemented it on our robot platform GripSee [1]. A novel object is placed into the robot s gripper in order to define a default orientation and a default grip. The robot then places the object on a turning table and builds up a visual representation that consists of a collection of graphs, labeled with multiscale edges. A user interface that ....

....of control must be left to the user. What is required may be called semi autonomy: a system (typically a robot) that can interact with the environment must dispose of a repertoire of skills that are carried out autonomously, but the actual control of behavior must be left to a human operator [1]. The knowledge dealt with here is not sophisticated high level knowledge about complicate interactions between things in the real world, because there is currently very little chance to acquire such knowledge without extensive programming. Instead we concentrate on modeling simple knowledge ....

[Article contains additional citation context not shown here]

M. Becker, E. Kefalea, E. Mael, C. von der Malsburg, M. Pagel, J. Triesch, J. C. Vorbruggen, R. P. Wurtz, and S. Zadel. GripSee: A gesture-controlled robot for object perception and manipulation. Autonomous Robots, 6(2), 1999. In press.

....of it on a table. The operator selects the object by pointing to it with a gesture indicating the grasping direction. To achieve this behavior, a sequence of other skills is necessary, such as hand tracking, hand posture analysis, grip planning, trajectory planning and grip execution discussed in [1]. In this procedure, hand tracking and posture analysis demonstrate our approach to human robot interaction. They are used to select an object and a grasp direction, but they are not needed to find the object. Thus, an alternative behavior could be started with the robot selecting one of the ....

M. Becker, E. Kefalea, E. Mael, C. von der Malsburg, M. Pagel, J. Triesch, J. C. Vorbruggen, R. P. Wurtz, and S. Zadel. GripSee: A gesturecontrolled robot for object perception and manipulation. Autonomous Robots, 1998. Submitted.

....measure for the quality of hand localization. Additional functionality, which is currently being worked on, includes the construction and incorporation of tactile sensors and the semi autonomous learning of new objects. More detailed information can be found on our website and in the articles [1, 2]. This work is funded by grants from the DFG (Graduiertenkolleg KOGNET) and the German Federal Minister for Education and Research (01 IN 504 E9) We thank Michael Potzsch, Michael Rinne, Bernd Fritzke, Herbert Jan en, Percy Dahm, Rainer Menzner, Thomas Bergener, and Carsten Bruckhoff for ....

M. Becker, E. Kefalea, E. Mael, C. von der Malsburg, M. Pagel, J. Triesch, J. C. Vorbruggen, R. P. Wurtz, and S. Zadel. GripSee: A gesture-controlled robot for object perception and manipulation. Autonomous Robots, 1998. Submitted.

No context found.

Becker M, Kefalea E, Mael E et al. 1999. `GripSee: A Gesture-controlled Robot for Object Perception and Manipulation'. Autom Robots, 6:203--221.

No context found.

M. Becker, E. Kefalea, E. Mael, C. V. D. Malsburg, M. Pagel, J. Triesch, J. C. Vorbruggen, R. P. Wurtz, and S. Zadel, "GripSee: A Gesture-controlled Robot for Object Perception and Manipulation," Autonomous Robots no. 6, pp. 203-221, 1999.

No context found.

M. Becker, E. Kefalea, E. Mael, C. V. D. Malsburg, M. Pagel, J. Triesch, J. C. Vorbruggen, R. P. Wurtz, and S. Zadel, "GripSee: A Gesturecontrolled Robot for Object Perception and Manipulation," Autonomous Robots no. 6, pp. 203-221, 1999.

No context found.

M. Becker, E. Kefalea, E. Mael, C. V. D. Malsburg, M. Pagel, J. Triesch, J. C. Vorbruggen, R. P. Wurtz, and S. Zadel. Gripsee: A gesture-controlled robot for object perception and manipulation. Autonomous Robots, (6):203--211, 1999.

No context found.

M. Becker, E. Kefalea, E. Ma el, C. von der Malsburg, M. Pagel, J. Triesch, J. C. Vorbr uggen, R. P. W urtz, and S. Zadel, "GripSee: A gesture-controlled robot for object perception and manipulation," Autonomous Robots, vol. 6, pp. 203--221, 1999.

No context found.

M. Becker, E. Kefalea, E. Mal, C. von der Malsburg, M. Pagel, J. Triesch, J.C. Vorbrggen, R.P. Wrtz, and S. Zadel. Gripsee: A gesture-controlled robot for object perception and manipulation. Autonomous Robots, 6:203{ 21, 1999.

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