KPRNet: Improving projection-based LiDAR semantic segmentation

Abstract

Semantic segmentation is an important component in the perception systems of autonomous vehicles. In this work, we adopt recent advances in both image and point cloud segmentation to achieve a better accuracy in the task of segmenting LiDAR scans. KPRNet improves the convolutional neural network architecture of 2D projection methods and utilizes KPConv to replace the commonly used post-processing techniques with a learnable point-wise component which allows us to obtain more accurate 3D labels. With these improvements our model outperforms the current best method on the SemanticKITTI benchmark, reaching an mIoU of 63.1.

"ECCV 2020. Code and pre-trained models at https://github.com/DeyvidKochanov-TomTom/kprnet"

References in corpus (2)

• Rethinking Atrous Convolution for Semantic Image Segmentation
• High-Resolution Representations for Labeling Pixels and Regions

Cited by in corpus (7)

• Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR Segmentation
• (AF)2-S3Net: Attentive Feature Fusion with Adaptive Feature Selection for Sparse Semantic Segmentation Network
• AMVNet: Assertion-based Multi-View Fusion Network for LiDAR Semantic Segmentation
• Panoptic-PolarNet: Proposal-free LiDAR Point Cloud Panoptic Segmentation
• S3Net: 3D LiDAR Sparse Semantic Segmentation Network
• Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR-based Perception
• Lite-HDSeg: LiDAR Semantic Segmentation Using Lite Harmonic Dense Convolutions