Enhancing Point Cloud Video Representation Learning through Partial Differential Equation Modeling

Modeling spatial-temporal correlations in point cloud videos as solvable partial differential equations (PDEs) to enhance representation learning and improve performance on downstream tasks.