Hiroaki Gomi

A Kendama learning robot based on bi-directional theory

A general theory of movement-pattern perception based on bi-directional theory for sensory-motor integration can be used for motion capture and learning by watching in robotics. We demonstrate our methods using the game of Kendama, executed by the SARCOS Dextrous Slave Arm, which has a very similar kinematic structure to the human arm. Three ingredients have to be integrated for the successful execution of this task. The ingredients are (1) to extract via-points from a human movement trajectory using a forward-inverse relaxation model, (2) to treat via-points as a control variable while reconstructing the desired trajectory from all the via-points, and (3) to modify the via-points for successful execution. In order to test the validity of the via-point representation, we utilized a numerical model of the SARCOS arm, and examined the behavior of the system under several conditions. Copyright 1996 Elsevier Science Ltd.

Large-Field Visual Motion Directly Induces an Involuntary Rapid Manual Following Response

Recent neuroscience studies have been concerned with how aimed movements are generated on the basis of target localization. However, visual information from the surroundings as well as from the target can influence arm motor control, in a manner similar to known effects in postural and ocular motor control. Here, we show an ultra-fast manual motor response directly induced by a large-field visual motion. This rapid response aided reaction when the subject moved his hand in the direction of visual motion, suggesting assistive visually evoked manual control during postural movement. The latency of muscle activity generating this response was as short as that of the ocular following responses to the visual motion. Abrupt visual motion entrained arm movement without affecting perceptual target localization, and the degrees of motion coherence and speed of the visual stimulus modulated this arm response. This visuomotor behavior was still observed when the visual motion was confined to the “follow-through” phase of a hitting movement, in which no target existed. An analysis of the arm movements suggests that the hitting follow through made by the subject is not a part of a reaching movement. Moreover, the arm response was systematically modulated by hand bias forces, suggesting that it results from a reflexive control mechanism. We therefore propose that its mechanism is radically distinct from motor control for aimed movements to a target. Rather, in an analogy with reflexive eye movement stabilizing a retinal image, we consider that this mechanism regulates arm movements in parallel with voluntary motor control.

Space and Time in Perception and Action: The utility of visual motion for goal-directed reaching

Visual information is crucial for goal-directed reaching. Recently a number of studies have shown that motion in particular is an important source of information for the visuomotor system. For example, when reaching for a stationary object, nearby visual movement even when irrelevant to the object or task can influence the trajectory of the hand. Although it is clear that various kinds of visual motion can influence goal-directed reaching movements, it is less clear how or why they do so. In this chapter, we consider whether the influence of motion on reaching is unique compared to its influence on other forms of visually guided behavior. We also address how motion is coded by the visuomotor system and whether there is one motion processing system that underlies both perception and visually guided reaching. Ultimately, visual motion may operate on a number of levels, influencing goal-directed reaching through more than one mechanism, some of which may actually be beneficial for accurate behavior.

A Kendama learning robot based on a dynamic optimization theory

A general theory of movement pattern perception based on a dynamic optimization theory can be used for motion capture and learning by watching in robotics. We exemplify our methods for the game of Kendama, executed by the SARCOS Dextrous Slave Arm, which has exactly the same kinematic structure as a human arm. Three ingredients have to be integrated for the successful execution of this task. The ingredients were (1) to extract via-points from a human movement trajectory using a forward-inverse relaxation model, (2) to treat via-points as a control variable while reconstructing the desired trajectory from all the via-points, and (3) to modify the via-points for successful execution.