Abstract:In this paper, we introduce a Point Recurrent Neural Network (PointRNN) for moving point cloud processing. At each time step, PointRNN takes point coordinates $\boldsymbol{P} \in \mathbb{R}^{n \times 3}$ and point features $\boldsymbol{X} \in \mathbb{R}^{n \times d}$ as input ($n$ and $d$ denote the number of points and the number of feature channels, respectively). The state of PointRNN is composed of point coordinates $\boldsymbol{P}$ and point states $\boldsymbol{S} \in \mathbb{R}^{n \times d'}$ ($d'$ denotes the number of state channels). Similarly, the output of PointRNN is composed of $\boldsymbol{P}$ and new point features $\boldsymbol{Y} \in \mathbb{R}^{n \times d''}$ ($d''$ denotes the number of new feature channels). Since point clouds are orderless, point features and states from two time steps can not be directly operated. Therefore, a point-based spatiotemporally-local correlation is adopted to aggregate point features and states according to point coordinates. We further propose two variants of PointRNN, i.e., Point Gated Recurrent Unit (PointGRU) and Point Long Short-Term Memory (PointLSTM). We apply PointRNN, PointGRU and PointLSTM to moving point cloud prediction, which aims to predict the future trajectories of points in a set given their history movements. Experimental results show that PointRNN, PointGRU and PointLSTM are able to produce correct predictions on both synthetic and real-world datasets, demonstrating their ability to model point cloud sequences. The code has been released at \url{<a class="link-external link-https" href="https://github.com/hehefan/PointRNN" rel="external noopener nofollow">this https URL</a>}.

PointRNN: Point Recurrent Neural Network for Moving Point Cloud Processing

RPP-Net: Rigid Constrained Point Cloud Prediction Network

RFNet - Recurrent Forward Network for Dense Point Cloud Completion.

Spiking PointNet: Spiking Neural Networks for Point Clouds

RCP: Recurrent Closest Point for Point Cloud

PointMotionNet: Point-Wise Motion Learning for Large-Scale LiDAR Point Clouds Sequences

Point2Point : A Framework for Efficient Deep Learning on Hilbert sorted Point Clouds with applications in Spatio-Temporal Occupancy Prediction

MoNet: Motion-Based Point Cloud Prediction Network

Modeling The Intensity Function Of Point Process Via Recurrent Neural Networks

3D Point Cloud Target Detection Method Based on R-PointGNN

Learning Adaptive Receptive Fields for Point Clouds.

PointRCNN: 3D Object Proposal Generation and Detection from Point Cloud

TriNN: A Concise, Lightweight, and Fast Global Triangulation GNN for Point Cloud

Spiking PointCNN: an Efficient Converted Spiking Neural Network under a Flexible Framework

DensePoint: Learning Densely Contextual Representation for Efficient Point Cloud Processing

PointHGN: Point Heterogeneous Graph Neural Network for Point Cloud Learning

Parameter is Not All You Need: Starting from Non-Parametric Networks for 3D Point Cloud Analysis

PointRWKV: Efficient RWKV-Like Model for Hierarchical Point Cloud Learning

RCP: Recurrent Closest Point for Scene Flow Estimation on 3D Point Clouds

NPTC-net: Narrow-Band Parallel Transport Convolutional Neural Network on Point Clouds