GPU-Accelerated Image-Based Feature Extraction and q-State Potts Model Simulation for 2D Active Materials in Biological Systems and Statistical Mechanics

This talk presents a novel integration of image analysis techniques for biological systems with a GPU-accelerated cluster algorithm designed to simulate 2D active materials using the q-state Potts model. The image analysis approach enables efficient preprocessing, feature extraction, and dynamic analysis of morphological growth patterns, including key metrics such as circularity and fractal dimensions. In parallel, the cluster algorithm harnesses GPU processing to simulate phase transitions and critical phenomena in active materials, providing real-time insights into system behavior at critical points. By combining these techniques, the workflow offers a computational toolkit for advancing both biological and physical research, allowing for enhanced image-based feature extraction and predictive modeling of critical phenomena. This talk will detail the underlying methodologies, code workflows, and real-world applications, demonstrating how this integrated approach streamlines research in both fields.