Abstract: The effective representation and feature extraction of 3-D scenes from sparse and unstructured point clouds pose a significant challenge in 3-D object detection. In this article, we propose ...