DEL: Discrete Element Learner for Learning 3D Particle Dynamics with Neural Rendering
This addresses the challenge of simulating particle dynamics without per-particle correspondences for applications in materials science and physics, representing an incremental improvement by combining existing methods.
The paper tackles the problem of learning 3D particle dynamics from 2D images by integrating learnable graph kernels into the Discrete Element Analysis framework to mitigate 2D-to-3D uncertainty, resulting in outperforming other learned simulators with robustness to different renderers, fewer training samples, and fewer camera views.
Learning-based simulators show great potential for simulating particle dynamics when 3D groundtruth is available, but per-particle correspondences are not always accessible. The development of neural rendering presents a new solution to this field to learn 3D dynamics from 2D images by inverse rendering. However, existing approaches still suffer from ill-posed natures resulting from the 2D to 3D uncertainty, for example, specific 2D images can correspond with various 3D particle distributions. To mitigate such uncertainty, we consider a conventional, mechanically interpretable framework as the physical priors and extend it to a learning-based version. In brief, we incorporate the learnable graph kernels into the classic Discrete Element Analysis (DEA) framework to implement a novel mechanics-integrated learning system. In this case, the graph network kernels are only used for approximating some specific mechanical operators in the DEA framework rather than the whole dynamics mapping. By integrating the strong physics priors, our methods can effectively learn the dynamics of various materials from the partial 2D observations in a unified manner. Experiments show that our approach outperforms other learned simulators by a large margin in this context and is robust to different renderers, fewer training samples, and fewer camera views.