Segment 3D scene with only class existence tags by using scene primitives with Densify Your Labels
Segment 3D scene with only class existence tags by using scene primitives with Densify Your Labels
Densify Your Labels: Unsupervised Clustering with Bipartite Matching for Weakly Supervised Point Cloud Segmentation
arXiv paper abstract https://arxiv.org/abs/2312.06799
arXiv PDF paper https://arxiv.org/pdf/2312.06799.pdf
Project page https://densify-your-labels.github.io
... propose a weakly supervised semantic segmentation method for point clouds that predicts "per-point" labels from just "whole-scene" annotations while achieving the performance of recent fully supervised approaches.
... core idea is to propagate the scene-level labels to each point in the point cloud by creating pseudo labels in a conservative way.
... over-segment point cloud features via unsupervised clustering and associate scene-level labels with clusters through bipartite matching, thus propagating scene labels only to the most relevant clusters, leaving the rest to be guided solely via unsupervised clustering.
... empirically demonstrate that over-segmentation and bipartite assignment plays a crucial role.
... method ... outperforming state of the art, and demonstrate ... can achieve results comparable to fully supervised methods.
Please like and share this post if you enjoyed it using the buttons at the bottom!
Stay up to date. Subscribe to my posts https://morrislee1234.wixsite.com/website/contact
Web site with my other posts by category https://morrislee1234.wixsite.com/website
Kommentare