LiDARNet: A Boundary-Aware Domain Adaptation Model for Point Cloud Semantic Segmentation Academic Article uri icon

abstract

  • We present a boundary-aware domain adaptation model for LiDAR scan full-scene semantic segmentation (LiDARNet). Our model can extract both the domain private features and the domain shared features with a two-branch structure. We embedded Gated-SCNN into the segmentor component of LiDARNet to learn boundary information while learning to predict full-scene semantic segmentation labels. Moreover, we further reduce the domain gap by inducing the model to learn a mapping between two domains using the domain shared and private features. Additionally, we introduce a new dataset (SemanticUSLfootnote{The access address of SemanticUSL:url{https://unmannedlab.github.io/research/SemanticUSL}}) for domain adaptation for LiDAR point cloud semantic segmentation. The dataset has the same data format and ontology as SemanticKITTI. We conducted experiments on real-world datasets SemanticKITTI, SemanticPOSS, and SemanticUSL, which have differences in channel distributions, reflectivity distributions, diversity of scenes, and sensors setup. Using our approach, we can get a single projection-based LiDAR full-scene semantic segmentation model working on both domains. Our model can keep almost the same performance on the source domain after adaptation and get an 8\%-22\% mIoU performance increase in the target domain.

author list (cited authors)

  • Jiang, P., & Saripalli, S.

publication date

  • January 1, 2020 11:11 AM