aiqbench analysis
This research introduces SLUM-i, a semi-supervised learning framework designed to improve the mapping of informal settlements using remote-sensing imagery with limited labeled data. By leveraging vision foundation models and adaptive thresholding techniques, the method achieves significant performance gains across seven diverse cities while adding no inference overhead. The accompanying benchmark quantifies cross-city heterogeneity in urban morphology, offering a robust resource for developing more equitable geospatial AI systems.
SLUM-i: Semi-supervised Learning for Urban Mapping of Informal Settlements and Data Quality Benchmarking
arXiv:2602.04525v3 Announce Type: replace-cross Abstract: Very-high-resolution remote-sensing imagery provides a scalable basis for delineating informal settlements, but sparse annotations, severe imbalance between informal-settlement and background pixels, and cross-city heterogeneity in urban morphology and image--mask correspondence complicate model development. We present SLUM-i, a semi-supervised semantic segmentation framework together with a geographically diverse seven-city Earth observation benchmark spanning three continents. The benchmark combines a newly annotated Lahore dataset and companion Karachi and Mumbai datasets with four publicly released city datasets, totaling 14,458 RGB image--mask tiles. We quantify cross-city heterogeneity using class composition, boundary morphology, grayscale separability, correspondence between mask boundaries and image edges, and divergence between measured feature distributions. For label-efficient mapping, SLUM-i combines representation-guided unlabeled-pool curation, using embeddings from a vision foundation model (DINOv2-Small) to remove the least-similar tiles, and Class-Aware Adaptive Thresholding, which adapts pseudo-label acceptance by class through global mean-confidence and per-class mean-softmax exponential moving averages. Experiments at 10%, 20%, and 30% label budgets, using convolutional and vision-transformer backbones and five random seeds, demonstrate improvements over the corresponding UniMatch baselines in multiple city--budget settings, reaching +5.9 percentage points in mean intersection-over-union. At the 30% budget, the ResNet-101 configuration matches or exceeds its corresponding fully labeled supervised baseline in four of seven cities. Both components operate only during training and add no inference overhead.
Abstract:Very-high-resolution remote-sensing imagery provides a scalable basis for delineating informal settlements, but sparse annotations, severe imbalance between informal-settlement and background pixels, and cross-city heterogeneity in urban morphology and image--mask correspondence complicate model development. We present SLUM-i, a semi-supervised semantic segmentation framework together with a geographically diverse seven-city Earth observation benchmark spanning three continents. The benchmark combines a newly annotated Lahore dataset and companion Karachi and Mumbai datasets with four publicly released city datasets, totaling 14,458 RGB image--mask tiles. We quantify cross-city heterogeneity using class composition, boundary morphology, grayscale separability, correspondence between mask boundaries and image edges, and divergence between measured feature distributions. For label-efficient mapping, SLUM-i combines representation-guided unlabeled-pool curation, using embeddings from a vision foundation model (DINOv2-Small) to remove the least-similar tiles, and Class-Aware Adaptive Thresholding, which adapts pseudo-label acceptance by class through global mean-confidence and per-class mean-softmax exponential moving averages. Experiments at 10%, 20%, and 30% label budgets, using convolutional and vision-transformer backbones and five random seeds, demonstrate improvements over the corresponding UniMatch baselines in multiple city--budget settings, reaching +5.9 percentage points in mean intersection-over-union. At the 30% budget, the ResNet-101 configuration matches or exceeds its corresponding fully labeled supervised baseline in four of seven cities. Both components operate only during training and add no inference overhead.
Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
arXiv:2602.04525 [cs.CV]
(or arXiv:2602.04525v3 [cs.CV] for this version)
https://doi.org/10.48550/arXiv.2602.04525 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 307 words
Related coverage
SLUM-i: Semi-supervised Learning for Urban Mapping of Informal Settlements and Data Quality Benchmarking
arXiv · Jun 16, 2026
DeepC4: Deep Conditional Census-Constrained Clustering for Large-scale Multitask Spatial Disaggregation of Urban Morphology
arXiv · Sep 28, 2026
Dharma vs Adharma - The Perplexity of Being Alive
hillpost.in · Aug 25, 2026
Monitoring Post-Disaster Urban Recovery Using High-Resolution SAR Time Series and Unsupervised Learning: Evidence from the 2023 T\"urkiye-Syria Earthquake
arXiv · Jul 28, 2026