GFNet: Geometric Flow Network for 3D Point Cloud Semantic Segmentation
Abstract: Point cloud semantic segmentation from projected views, such as range-view (RV) and bird's-eye-view (BEV), has been intensively investigated. Different views capture different information of point clouds and thus are complementary to each other. However, recent projection-based methods for point cloud semantic segmentation usually utilize a vanilla late fusion strategy for the predictions of different views, failing to explore the complementary information from a geometric perspective during the representation learning. In this paper, we introduce a geometric flow network (GFNet) to explore the geometric correspondence between different views in an align-before-fuse manner. Specifically, we devise a novel geometric flow module (GFM) to bidirectionally align and propagate the complementary information across different views according to geometric relationships under the end-to-end learning scheme. We perform extensive experiments on two widely used benchmark datasets, SemanticKITTI and nuScenes, to demonstrate the effectiveness of our GFNet for project-based point cloud semantic segmentation. Concretely, GFNet not only significantly boosts the performance of each individual view but also achieves state-of-the-art results over all existing projection-based models. Code is available at \url{https://github.com/haibo-qiu/GFNet}.
Submission Length: Regular submission (no more than 12 pages of main content)
Changes Since Last Submission: We appreciate all valuable comments/suggestions from AE the reviewers. Based on their concerns, we have carefully modified our paper in this camera-ready version as follows:
1. We have incorporated our answers regarding 3D object detection and indoor scenes as limitations in the last paragraph of Sec.5 (Conclusion).
2. We have added Sec.6 (Acknowledgement)
3. Previous highlight color has been removed.
4. Information about authors and the OpenReview link has been added.
Code: https://github.com/haibo-qiu/GFNet
Assigned Action Editor: ~Mathieu_Salzmann1
License: Creative Commons Attribution 4.0 International (CC BY 4.0)
Submission Number: 238
Loading