Large-Scale Semantic Scene Understanding with Cross-Correction Representation

Abstract

Real-time large-scale point cloud segmentation is an important but challenging task for practical applications such as remote sensing and robotics. Existing real-time methods have achieved acceptable performance by aggregating local information. However, most of them only exploit local spatial geometric or semantic information dependently, few considering the complementarity of both. In this paper, we propose a model named Spatial–Semantic Incorporation Network (SSI-Net) for real-time large-scale point cloud segmentation. A Spatial-Semantic Cross-correction (SSC) module is introduced in SSI-Net as a basic unit. High-quality contextual features can be learned through SSC by correcting and updating high-level semantic information using spatial geometric cues and vice versa. Adopting the plug-and-play SSC module, we design SSI-Net as an encoder–decoder architecture. To ensure efficiency, it also adopts a random sample-based hierarchical network structure. Extensive experiments on several prevalent indoor and outdoor datasets for point cloud semantic segmentation demonstrate that the proposed approach can achieve state-of-the-art performance.