@article{ART003258727},
author={Yoonsik Shim},
title={A Framework for Learning and Reuse of Motor Patterns Emerged from Embodied Chaotic Search},
journal={Journal of The Korea Society of Computer and Information},
issn={1598-849X},
year={2025},
volume={30},
number={10},
pages={61-68}
TY - JOUR
AU - Yoonsik Shim
TI - A Framework for Learning and Reuse of Motor Patterns Emerged from Embodied Chaotic Search
JO - Journal of The Korea Society of Computer and Information
PY - 2025
VL - 30
IS - 10
PB - The Korean Society Of Computer And Information
SP - 61
EP - 68
SN - 1598-849X
AB - We present an improved framework for stabilizing and reusing self-organized motor patterns discovered through chaotic search in embodied neuromechanical systems. Based on the Embodied Chaotic Search (ECS) paradigm— which enables the emergent generation of periodic motor patterns via neuro-body-environment interactions without episode-based machine learning—our approach addresses the inherent limitation of pattern instability. The proposed method introduces a two-stage process that decouples search and learning. Upon detecting a promising motor pattern, external sensory feedback is suppressed to freeze the system’s dynamics, enabling stable learning of inter-CPG coupling weights through adaptive synchronization. The learned weights are stored and can be reloaded to reliably reproduce the pattern. Phase space deformation before and after learning is visualized, demonstrating the convergence of multiple transient attractors into a single stable one. To reduce complexity, we replace the rotation-center correction of prior adaptive synchronization algorithms with a bias-based learning mechanism compatible with standard neural models. This framework improves the acquisition and memory stability of the ECS system and highlights the potential of leveraging these patterns as initial policies for conventional machine learning, as well as foundational motor primitives for high-level behavioral synthesis in robotic systems.
KW - Embodied AI;Self organization;Chaotic exploration;Central pattern generator;;Adaptive synchronization;Motor primitive;Neuromechanical system
DO -
UR -
ER -
Yoonsik Shim. (2025). A Framework for Learning and Reuse of Motor Patterns Emerged from Embodied Chaotic Search. Journal of The Korea Society of Computer and Information, 30(10), 61-68.
Yoonsik Shim. 2025, "A Framework for Learning and Reuse of Motor Patterns Emerged from Embodied Chaotic Search", Journal of The Korea Society of Computer and Information, vol.30, no.10 pp.61-68.
Yoonsik Shim "A Framework for Learning and Reuse of Motor Patterns Emerged from Embodied Chaotic Search" Journal of The Korea Society of Computer and Information 30.10 pp.61-68 (2025) : 61.
Yoonsik Shim. A Framework for Learning and Reuse of Motor Patterns Emerged from Embodied Chaotic Search. 2025; 30(10), 61-68.
Yoonsik Shim. "A Framework for Learning and Reuse of Motor Patterns Emerged from Embodied Chaotic Search" Journal of The Korea Society of Computer and Information 30, no.10 (2025) : 61-68.
Yoonsik Shim. A Framework for Learning and Reuse of Motor Patterns Emerged from Embodied Chaotic Search. Journal of The Korea Society of Computer and Information, 30(10), 61-68.
Yoonsik Shim. A Framework for Learning and Reuse of Motor Patterns Emerged from Embodied Chaotic Search. Journal of The Korea Society of Computer and Information. 2025; 30(10) 61-68.
Yoonsik Shim. A Framework for Learning and Reuse of Motor Patterns Emerged from Embodied Chaotic Search. 2025; 30(10), 61-68.
Yoonsik Shim. "A Framework for Learning and Reuse of Motor Patterns Emerged from Embodied Chaotic Search" Journal of The Korea Society of Computer and Information 30, no.10 (2025) : 61-68.
