Improving Lidar-Based Semantic Segmentation of Top-View Grid Maps by Learning Features in Complementary Representations
This work addresses the challenge of accurate semantic segmentation in autonomous driving systems, but it is incremental as it builds on existing methods by optimizing feature fusion.
The paper tackles the problem of semantic segmentation of top-view grid maps from sparse LiDAR data for autonomous driving by fusing learned features from complementary 2D representations, achieving improved performance as evaluated on the SemanticKITTI dataset with over 23,000 annotated measurements.
In this paper we introduce a novel way to predict semantic information from sparse, single-shot LiDAR measurements in the context of autonomous driving. In particular, we fuse learned features from complementary representations. The approach is aimed specifically at improving the semantic segmentation of top-view grid maps. Towards this goal the 3D LiDAR point cloud is projected onto two orthogonal 2D representations. For each representation a tailored deep learning architecture is developed to effectively extract semantic information which are fused by a superordinate deep neural network. The contribution of this work is threefold: (1) We examine different stages within the segmentation network for fusion. (2) We quantify the impact of embedding different features. (3) We use the findings of this survey to design a tailored deep neural network architecture leveraging respective advantages of different representations. Our method is evaluated using the SemanticKITTI dataset which provides a point-wise semantic annotation of more than 23.000 LiDAR measurements.