Special Session 2
Training in Robotics for Development of Cognition (RobotDoC)
Abstract
This special session will present recent advances in robotic modelling of cognitive development as well as behavioural studies with implications for robotics. It summarises the main results of the European Marie Curie doctoral training network RobotDoC. The special session is part of the co-located RobotDoC Conference that will be held in Osaka on August 16-18.
Research in the RobotDoC training network specifically focuses on interdisciplinary approaches to developmental cognitive robotics (Asada et al., 2001; Cangelosi & Schlesinger, to appear). In particular, we present here a syncretic overview of new models and experimental results in: (i) Dimensionality reduction for efficient motor control and new hardware implementations for sensorimotor systems, (ii) using sensorimotor representations for developing adaptive robots and for grounding abstract knowledge, (iii) social attention and visual action prediction for interactive robots, and (iv) developmental approaches to social interaction and robot feedback in human-robot interaction. Moreover, we will discuss insights and lessons learned on training methodology for researchers in developmental robotics.
[1] M. Asada, K.F. MacDorman, H. Ishiguro and Y. Kuniyoshi, “Cognitive
developmental robotics as a new paradigm for the design of humanoid
robots,” Robotics and Autonomous Systems, 37(2), 185-193, 2001.
[2] A. Cangelosi and M. Schlesinger, “An introduction to developmental
robotics,” MIT Press, Cambridge, MA, to appear.