SLUM-i: Semi-supervised Learning for Urban Mapping of Informal Settlements and Data Quality Benchmarking
Researchers have developed a new semi-supervised learning framework called SLUM-i to improve the mapping of informal urban settlements. This method addresses challenges like limited annotations and data quality issues, particularly in cities like Lahore, Karachi, and Mumbai. The framework incorporates a Class-Aware Adaptive Thresholding mechanism to prevent minority class suppression and a DINOv2-based filter to remove irrelevant data, demonstrating significant improvements in segmentation accuracy over existing methods. AI
IMPACT This research offers a novel approach to semi-supervised learning for urban mapping, potentially improving data quality and accessibility for informal settlements.