BelHouse3D: A Dataset for 3D Indoor Scene Point Clouds

The BelHouse3D dataset is a synthetic point cloud dataset for 3D indoor scene semantic segmentation. It is constructed using real-world references from 32 houses in Belgium, ensuring that the synthetic data closely aligns with real-world conditions. Additionally, it includes a test set with data occlusion to simulate out-of-distribution (OOD) scenarios, reflecting the occlusions commonly encountered in real-world point clouds. The dataset is used to benchmark both fully supervised and few-shot learning (FSL) segmentation models.