Panoptic-PolarNet: Proposal-free LiDAR Point Cloud Panoptic Segmentation
arXiv:2103.14962
Abstract
Panoptic segmentation presents a new challenge in exploiting the merits of both detection and segmentation, with the aim of unifying instance segmentation and semantic segmentation in a single framework. However, an efficient solution for panoptic segmentation in the emerging domain of LiDAR point cloud is still an open research problem and is very much under-explored. In this paper, we present a fast and robust LiDAR point cloud panoptic segmentation framework, referred to as Panoptic-PolarNet. We learn both semantic segmentation and class-agnostic instance clustering in a single inference network using a polar Bird's Eye View (BEV) representation, enabling us to circumvent the issue of occlusion among instances in urban street scenes. To improve our network's learnability, we also propose an adapted instance augmentation technique and a novel adversarial point cloud pruning method. Our experiments show that Panoptic-PolarNet outperforms the baseline methods on SemanticKITTI and nuScenes datasets with an almost real-time inference speed. Panoptic-PolarNet achieved 54.1% PQ in the public SemanticKITTI panoptic segmentation leaderboard and leading performance for the validation set of nuScenes.
Accepted by CVPR 2021
References in corpus (12)
- YOLOv4: Optimal Speed and Accuracy of Object Detection
- DropBlock: A regularization method for convolutional networks
- Semantic Instance Segmentation with a Discriminative Loss Function
- A2D2: Audi Autonomous Driving Dataset
- Scalability in Perception for Autonomous Driving: Waymo Open Dataset
- DeeperLab: Single-Shot Image Parser
- SalsaNext: Fast, Uncertainty-aware Semantic Segmentation of LiDAR Point Clouds for Autonomous Driving
- KPRNet: Improving projection-based LiDAR semantic segmentation
- The Lovász-Softmax loss: A tractable surrogate for the optimization of the intersection-over-union measure in neural networks
- MOPT: Multi-Object Panoptic Tracking
- Quantifying Data Augmentation for LiDAR based 3D Object Detection
- Self-Prediction for Joint Instance and Semantic Segmentation of Point Clouds