Efficient Instance Motion-Aware Point Cloud Scene Prediction

Point cloud prediction (PCP) aims to forecast future 3D point clouds of scenes by leveraging sequential historical LiDAR scans, offering a promising avenue to enhance the perceptual capabilities of autonomous systems. However, existing methods mostly adopt an end-to-end approach without explicitly modeling moving instances, limiting their effectiveness in dynamic real-world environments. In this paper, we propose IMPNet, a novel instance motion-aware network for future point cloud scene prediction. Unlike prior works, IMPNet explicitly incorporates motion and instance-level information to enhance PCP accuracy. Specifically, we extract appearance and motion features from range images and residual images using a dual-branch convolutional network and fuse them via a motion attention block. Our framework further integrates a motion head for identifying moving objects and an instance-assisted training strategy to improve instance-wise point cloud predictions. Extensive experiments on multiple datasets demonstrate that our proposed network achieves state-of-the-art (SOTA) performance in PCP with superior predictive accuracy and robust generalization across diverse driving scenarios. Our method has been released at https://github.com/nubot-nudt/IMPNet.

Authors

• Yiming Fang
• Xieyuanli Chen
• Neng Wang
• Kaihong Huang
• Huimin Lu 0002

Keywords

• Point cloud compression
• Training
• Accuracy
• Three-dimensional displays
• Limiting
• Laser radar
• Object segmentation
• Feature extraction
• Convolutional neural networks
• Object recognition
• Point Cloud
• Convolutional Network
• Motion Features
• Appearance Features
• 3D Point Cloud
• Motion Information
• Superior Prediction
• LiDAR Scans
• Extract Motion
• Spatiotemporal
• Time Step
• Superior Performance
• Spatial Information
• Feature Maps
• Bounding Box
• Temporal Information
• Attention Module
• Path Planning
• Instance Segmentation
• 2D Convolutional Network
• Simultaneous Localization And Mapping
• Chamfer Distance
• Superior Predictive Performance
• LiDAR Sensor
• Point Cloud Data
• Convolutional Long Short-term Memory
• Laser Pointer
• Point Cloud Generation
• Average Loss