PointSSC: A Cooperative Vehicle-Infrastructure Point Cloud Benchmark for Semantic Scene Completion

TL;DR

PointSSC introduces a new outdoor vehicle-infrastructure point cloud benchmark for semantic scene completion, featuring a novel annotation pipeline and a LiDAR-based model with a Spatial-Aware Transformer to advance real-world navigation tasks.

Contribution

We present the first cooperative vehicle-infrastructure point cloud benchmark with automated semantic annotation and a novel LiDAR-based model for semantic scene completion.

Findings

Benchmark facilitates progress in outdoor semantic point cloud completion.

Proposed model effectively combines global and local features for scene understanding.

Code and datasets are publicly available for research use.

Abstract

Semantic Scene Completion (SSC) aims to jointly generate space occupancies and semantic labels for complex 3D scenes. Most existing SSC models focus on volumetric representations, which are memory-inefficient for large outdoor spaces. Point clouds provide a lightweight alternative but existing benchmarks lack outdoor point cloud scenes with semantic labels. To address this, we introduce PointSSC, the first cooperative vehicle-infrastructure point cloud benchmark for semantic scene completion. These scenes exhibit long-range perception and minimal occlusion. We develop an automated annotation pipeline leveraging Semantic Segment Anything to efficiently assign semantics. To benchmark progress, we propose a LiDAR-based model with a Spatial-Aware Transformer for global and local feature extraction and a Completion and Segmentation Cooperative Module for joint completion and segmentation. PointSSC provides a challenging testbed to drive advances in semantic point cloud completion for real-world navigation. The code and datasets are available at https://github.com/yyxssm/PointSSC.