Coupled Oscillator
Coupled oscillators, systems of interacting oscillating elements, are studied to understand emergent collective behaviors and their applications in computation and robotics. Current research focuses on leveraging coupled oscillator models, such as the Kuramoto model, within neural network architectures (e.g., graph neural networks) and neuromorphic computing platforms to improve efficiency and address challenges like over-smoothing. These investigations are significant because they offer insights into complex dynamical systems and enable the development of novel computational tools with potential applications in areas ranging from signal processing and machine learning to underwater robotics and collective sensing.
Papers
Learning Emergent Gaits with Decentralized Phase Oscillators: on the role of Observations, Rewards, and Feedback
Jenny Zhang, Steve Heim, Se Hwan Jeon, Sangbae Kim
Training Coupled Phase Oscillators as a Neuromorphic Platform using Equilibrium Propagation
Qingshan Wang, Clara C. Wanjura, Florian Marquardt