We introduce a dual-head structure designed to jointly handle change classification and cross-visibility confidence estimation.
Chamelion is a novel approach for online change detection and long-term 3D map management, leveraging a dual-head network and composition-based augmentation for robust generalization across sensors and environments.
Training change detection requires multi-session data, which is costly and impractical to collect. To address this, we propose a single-session augmentation method that synthetically generates change pseudo-labels using only a single traversal of the environment.
We introduce a dual-head structure designed to jointly handle change classification and cross-visibility confidence estimation.
The class head predicts whether each point is a change or static, while the confidence head estimates the probability that the point is visible in both the map and scan.
We qualitatively evaluated the generalization performance across various environments.
Compare the prior map and current scans from Const-1F. The views share the same camera for inspecting low-dynamic (LD) changes.
● Prior map ● Current scan ● Negative change ● Positive change
Map & scan cross visibility: ● Low (0) ● High (1) ● Outside frame input
Drag to rotate · Scroll or pinch to zoom · Right-drag to pan. Browse 100 scans sampled across the full sequence. The final map change result stays fixed; Cross visibility shows map and scan confidence for the selected frame.