Object Dynamics Modeling with Hierarchical Point Cloud-based Representations

arXiv:2404.06044v1 Announce Type: new
Abstract: Modeling object dynamics with a neural network is an important problem with numerous applications. Most recent work has been based on graph neural networks. However, physics happens in 3D space, where geometric information potentially plays an important role in modeling physical phenomena. In this work, we propose a novel U-net architecture based on continuous point convolution which naturally embeds information from 3D coordinates and allows for multi-scale feature representations with established downsampling and upsampling procedures. …

abstract applications arxiv cloud cloud-based cs.cv dynamics graph graph neural networks hierarchical however information modeling network networks neural network neural networks object physics role space type work