Uncertainty-Aware Point-Cloud Semantic Segmentation for Unstructured Roads
Pengfei Liu, Guizhen Yu, Zhangyu Wang, Bin Zhou, Ruotong Ming, Chunhua Jin
Abstract
Semantic segmentation is one of the fundamental elements for achieving effective and safe autonomous driving. However, due to the irregular boundaries and variable illumination of unstructured roads, applying it in these scenarios is confronted with great challenges. To address these problems, a novel point-cloud semantic segmentation framework for unstructured roads is proposed. It contains three sections: spherical projection, an uncertainty-aware semantic segmentation network, and postprocessing. First, point cloud will be projected to the range image, which can be processed by the 2-D convolution network. Then, the uncertainty-aware semantic segmentation network is constructed. It consists of context-aware attention (CAA) module and direction attention up-sampling (DAU) module, which can improve the performance for the segmentation of unstructured roads. In addition, a Gaussian mixture model (GMM) is introduced at the end of the network to predict the result with uncertainty, indicating the confidence level of the output. Finally, the segmentation result is refined during the postprocessing to help filter the noise points. Experimental data from mine sites were collected to validate the performance for unstructured roads. In addition, the proposed method was evaluated on the public unstructured dataset RELLIS-3-D. The experiments show that the proposed architecture achieved 74.9% and 40.4% mIoU, which performs better than comparison methods. Additionally, the network is more robust to noisy data by achieving improvements of 4.6%–7.6% under different levels of noise data.