In this paper, we propose an end-to-end learning network to predict future frames in a point cloud sequence. As main novelty, an initial layer learns topological information of point clouds as geometric features, to form representative spatio-temporal neighborhoods. This module is followed by multiple Graph-RNN cells. Each cell learns points dynamics (i.e., RNN states) by processing each point jointly with the spatio-temporal neighbouring points. We tested the network performance with a MINST dataset of moving digits, a synthetic human bodies motions and JPEG dynamic bodies datasets. Simulation results demonstrate that our method outperforms baseline ones that neglect geometry features information.
@article{arxiv.2102.07482,
title = {Spatio-temporal Graph-RNN for Point Cloud Prediction},
author = {Pedro Gomes and Silvia Rossi and Laura Toni},
journal= {arXiv preprint arXiv:2102.07482},
year = {2021}
}