GAF-Net: Geometric Contextual Feature Aggregation and Adaptive Fusion for Large-Scale Point Cloud Semantic Segmentation

Large-scale point cloud semantic segmentation is a challenging task due to the complexity and diversity of real-world 3-D scenes. Most existing methods primarily rely on spatial coordinates to learn geometric representations without fully exploring local structural relationships. Additionally, the semantic gap between the encoder and decoder in segmentation networks is an important factor that constrains model performance. To address these challenges, we propose a novel network architecture called GAF-Net, which comprises a geometric contextual feature aggregation (GCFA) module and a multiscale feature adaptive fusion (MFAF) module. The GCFA module consists of three primary blocks: 1) a geometric edge representation (GER) block, designed to leverage spatial relative position and orientation information between the center point and its neighbors to capture detailed local geometric structural relations; 2) a point geometry prior (PGP) block, aimed at extracting explicit geometric priors for each point from raw point clouds. This block is lightweight and parameter-free; and 3) a geometry-aware attentive pooling (GAAP) block, which combines semantic features with learned geometric representations, enabling the learning and aggregation of informative local contextual features. Our proposed MFAF module integrates multiscale features by introducing an adaptive fusion approach. It effectively bridges the semantic gap between the encoder and decoder and mitigates the information loss caused by random sampling. Extensive experimental results on three large-scale benchmark datasets.