Machine Learning Assisted Characterization of Labyrinthine Pattern Transitions

Labyrinthine structures are ubiquitous emerging patterns in out-of-equilibrium nonlinear systems. They exhibit disordered stripe patterns with different orientations, sizes, and grain-boundary structures without well-defined order parameters, leading to difficulty in their systematic characterization. In this study, we present a comprehensive approach to characterizing labyrinthine structures [1]. We develop advanced machine learning based pattern recognition techniques to identify the types and locations of topological defects in magnetic labyrinthine patterns from experimental images [2]. Applying this method to single-crystal Bi-substituted Yttrium Iron Garnet films, we uncover a distinct morphological transition between two zero-field labyrinthine structures. Crucially, the pair distribution functions of the topological defects reveal subtle differences between labyrinthine structures which are beyond conventional characterization methods. By systematically analyzing the spatial correlations and geometric properties of these defects, we provide new insights into the athermal dynamics governing the observed morphological transitions. Our work demonstrates that machine learning based recognition techniques enable novel studies of rich and complex labyrinthine type structures universal to many pattern formation systems.

[1] K. Shimizu et al., arXiv:2311.10558 (2023).

[2] V. Y. Okubo et al., IEEE Access 12, 92419 (2024).