CiNet Monthly Seminar
August 2, 2018
16:00 ~ 17:00
CiNet 1F Conference Room
“What is really learned during sensorimotor learning: Lessons from relearning”
School of Psychological and Cognitive Sciences
Host : Atsushi Yokoi (Nishimoto Group)
Motor learning manifests itself as performance improvements as well as faster relearning when the task is encountered again (i.e., savings).
For sensorimotor adaptation where the motor system learns to counter external perturbations, the prevailing view is that savings is caused by the retrieval of an explicit movement strategy. Here I present a series of human experiments to show that savings can be elicited by implicit learning, which is likely mediated by Cerebellum. Moreover, the spatial generalization of savings is bounded, echoing early findings that implicit learning is directionally tuned. Further experiments also find that savings can be upregulated by saliency of motor errors. In sum, our studies reveal that long-term motor learning can be achieved by enhanced implicit learning to motor errors, independent from explicit awareness of action strategies. The distinction between procedural and declarative knowledge will be questioned and discussed.
For the second part of my talk, I will introduce two applied research projects. The first project is to combine machine learning algorithms with eye-tracking data for the early screening of autism. The second project is to use hand exoskeleton and virtual reality in stroke rehabilitation.