October 2, 2025Open Access

FloorSAM: SAM-Guided Floorplan Reconstruction with Semantic-Geometric Fusion

View Full Paper

Ask AI

Bookmark

FloorSAM improves floorplan reconstruction accuracy from point cloud data, enhancing navigation efficiency.
Performance testing on Giblayout and ISPRS datasets revealed higher accuracy and recall than traditional methods.
The framework utilizes SAM's zero-shot learning for robust room segmentation, addressing noise and data complexity.
Joint mask and point cloud analysis allows FloorSAM to recover topological relationships in reconstructed layouts.

Discussion

No takes yet. Share an insight, caveat, or question.

FloorSAM integrates point cloud data and Segment Anything Model for improved reconstruction accuracy, suggesting advancements in indoor navigation.

Ye et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark

FloorSAM integrates point cloud data and Segment Anything Model for improved reconstruction accuracy, suggesting advancements in indoor navigation.