RegionNet: Region-feature-enhanced 3D Scene Understanding Network with Dual Spatial-aware Discriminative Loss
Guanghui Zhang, Dongchen Zhu, Xiaoqing Ye, Wenjun Shi, Minghong Chen, Jiamao Li, Xiaolin Zhang
Abstract
Neural networks have recently achieved impressive success in semantic and instance segmentation on 2D images. However, their capabilities have not been fully explored to address semantic instance segmentation on unstructured 3D point cloud data. Digging into the regional feature representation to boost point cloud comprehension, we propose a region-feature-enhanced structure consisting of adaptive regional feature complementary (ARFC) module and affinity-based regional relational reasoning (AR3) module. The ARFC module aims to complement low-level features of sparse regions adaptively. The AR3 module emphasizes on mining the potential reasoning relationships between high-level features based on affinity. Both the ARFC and AR3 modules are plug-and-play. Besides, a novel dual spatial-aware discriminative loss is proposed to improve the discrimination of instance embedding. Our proposal-free point cloud instance segmentation network (RegionNet) equipped with the region-feature-enhanced structure and dual spatial-aware discriminative loss achieves state-of-the-art performance on S3DIS dataset and ScanNet-v2 dataset.
BibTeX
@inproceedings{iros2020_regionnetregionf,
title = {RegionNet: Region-feature-enhanced 3D Scene Understanding Network with Dual Spatial-aware Discriminative Loss},
author = {Guanghui Zhang and Dongchen Zhu and Xiaoqing Ye and Wenjun Shi and Minghong Chen and Jiamao Li and Xiaolin Zhang},
booktitle = {IROS 2020},
year = {2020}
}