RBMatch: Dual-Level Class Rebalancing for Semi-Supervised Building Footprint Extraction

Key points

• In remote sensing, severe foreground--background imbalance poses a particular challenge for self-training, as it can bias pseudo-label generation and the resulting unsupervised optimization toward the majority background class.1
• We show that addressing this imbalance at only one stage is insufficient: balancing pseudo-label selection alone does not prevent background bias from re-emerging during unsupervised loss optimization, a failure mode we term imbalance leak.1
• To address this issue, we propose RBMatch, a dual-level class-rebalancing framework that jointly regulates pseudo-label generation and unsupervised optimization.1
• Experiments on the WHU, INRIA, and Massachusetts building footprint datasets across labeled ratios of 1%--10% show that RBMatch consistently achieves the best building IoU and F1-score among the evaluated methods.1

Sources (1)

• [1]RBMatch: Dual-Level Class Rebalancing for Semi-Supervised Building Footprint Extraction
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 6, 03:40 AM

  Accurate building footprint extraction from high-resolution remote sensing imagery is essential for urban planning, disaster response, and environmental monitoring.

  In remote sensing, severe foreground--background imbalance poses a particular challenge for self-training, as it can bias pseudo-label generation and the resulting unsupervised optimization toward the majority background class.

Extractive summary: sentences quoted from the sources.