Does the invariance in multi-modalities represent the body scheme ?-a case study with vision and proprioception -
Does the invariance in multi-modalities represent the body scheme ?-a case study with vision and proprioception -
复制标题
多模态的不变性是否代表了身体方案？-视觉和本体感觉的案例研究-
DOI：
--
复制
发表时间：
2003
期刊：
影响因子：
--
通讯作者：
M. Asada
中科院分区：
文献类型：
--
作者：
Y. Yoshikawa;K. Hosoda;M. Asada
对环境和机器人本体本身变化的适应性从根本上取决于机器人本体表示，该表示通常由设计者给出，因此在许多情况下是固定的。为了使机器人适应其身体表示的变化，机器人应该已经获得了自己的身体表示本身。本文讨论了机器人如何能够构建这样的表示，即身体方案或身体图像从未解释的行的感觉信息。假设多传感器数据中的不变性代表身体，提出了一种交叉模态映射作为学习不变性的结构。初步的实验，学习代表身体表面的机器人之间的视觉和本体感觉的交叉模态映射进行和未来的问题进行了讨论。
Adaptability to the changes in the environment and the robot body itself fundamentally depends on the robot body representation, which is usually given by the designer and therefore fixed in many cases. In order for the robot to adapt its body representation to the changes, the robot should have acquired its own body representation by itself. This paper argues how the robot can construct such representation, that is, body scheme or body image from the uninterpreted row sensory information. Supposing that the invariance in multi sensory data represents the body, a cross modal map is proposed as the structure which learns the invariance. A preliminary experiment to learn to represent the body surfaces of the robot by the cross modal mapping between vision and proprioception is performed and future issues are discussed.