Alert button
Picture for Haowen Guo

Haowen Guo

Alert button

Bidirectional Feature Globalization for Few-shot Semantic Segmentation of 3D Point Cloud Scenes

Aug 17, 2022
Yongqiang Mao, Zonghao Guo, Xiaonan Lu, Zhiqiang Yuan, Haowen Guo

Figure 1 for Bidirectional Feature Globalization for Few-shot Semantic Segmentation of 3D Point Cloud Scenes
Figure 2 for Bidirectional Feature Globalization for Few-shot Semantic Segmentation of 3D Point Cloud Scenes
Figure 3 for Bidirectional Feature Globalization for Few-shot Semantic Segmentation of 3D Point Cloud Scenes
Figure 4 for Bidirectional Feature Globalization for Few-shot Semantic Segmentation of 3D Point Cloud Scenes

Few-shot segmentation of point cloud remains a challenging task, as there is no effective way to convert local point cloud information to global representation, which hinders the generalization ability of point features. In this study, we propose a bidirectional feature globalization (BFG) approach, which leverages the similarity measurement between point features and prototype vectors to embed global perception to local point features in a bidirectional fashion. With point-to-prototype globalization (Po2PrG), BFG aggregates local point features to prototypes according to similarity weights from dense point features to sparse prototypes. With prototype-to-point globalization (Pr2PoG), the global perception is embedded to local point features based on similarity weights from sparse prototypes to dense point features. The sparse prototypes of each class embedded with global perception are summarized to a single prototype for few-shot 3D segmentation based on the metric learning framework. Extensive experiments on S3DIS and ScanNet demonstrate that BFG significantly outperforms the state-of-the-art methods.

* Institutional error 
Viaarxiv icon