PolarStream: Streaming Lidar Object Detection and Segmentation with Polar Pillars
arXiv:2106.07545
Abstract
Recent works recognized lidars as an inherently streaming data source and showed that the end-to-end latency of lidar perception models can be reduced significantly by operating on wedge-shaped point cloud sectors rather then the full point cloud. However, due to use of cartesian coordinate systems these methods represent the sectors as rectangular regions, wasting memory and compute. In this work we propose using a polar coordinate system and make two key improvements on this design. First, we increase the spatial context by using multi-scale padding from neighboring sectors: preceding sector from the current scan and/or the following sector from the past scan. Second, we improve the core polar convolutional architecture by introducing feature undistortion and range stratified convolutions. Experimental results on the nuScenes dataset show significant improvements over other streaming based methods. We also achieve comparable results to existing non-streaming methods but with lower latencies. The code and pretrained models are available at \url{https://github.com/motional/polarstream}.
NeurIPS 2021; code and pretrained models available at https://github.com/motional/polarstream
References in corpus (11)
- Fast and Furious: Real Time End-to-End 3D Detection, Tracking and Motion Forecasting with a Single Convolutional Net
- IntentNet: Learning to Predict Intention from Raw Sensor Data
- HDNET: Exploiting HD Maps for 3D Object Detection
- Cylinder3D: An Effective 3D Framework for Driving-scene LiDAR Semantic Segmentation
- End-to-End Multi-View Fusion for 3D Object Detection in LiDAR Point Clouds
- PolarNet: An Improved Grid Representation for Online LiDAR Point Clouds Semantic Segmentation
- (AF)2-S3Net: Attentive Feature Fusion with Adaptive Feature Selection for Sparse Semantic Segmentation Network
- Panoptic-PolarNet: Proposal-free LiDAR Point Cloud Panoptic Segmentation
- StrObe: Streaming Object Detection from LiDAR Packets
- It's All Around You: Range-Guided Cylindrical Network for 3D Object Detection
- Streaming Object Detection for 3-D Point Clouds