Slum Detection and Density Mapping with AlphaEarth Foundations: A Representation Learning Evaluation Across 12 Global Cities

2026-05-11

AI Summary

This study addresses the challenges of weak cross-city generalization, absence of continuous density estimation, and poor global comparability in slum mapping by presenting the first validation of AlphaEarth Foundations—a globally consistent 64-dimensional land surface embedding—for the indirectly related task of slum identification. Leveraging data from 12 cities, the authors systematically evaluate its performance in both classification and sub-pixel density estimation using GRAM pseudo-labels, POI-derived auxiliary features, spatially blocked cross-validation, and multiple model baselines. Results demonstrate optimal within-city cross-year transfer (F1=0.616, R²=0.466), a significant improvement in density modeling from POI integration (ΔR²=+0.064), and strong spatial structural consistency in six-city full-area inference (SSIM=0.926), while also revealing city-scale representation drift.

Abstract

Pixel-level slum mapping has long been constrained by limited cross-city generalisation, the absence of continuous density estimation, and weak global comparability. AlphaEarth Foundations (AEF), a globally consistent 64-dimensional annual surface embedding at 10 m, offers a new analysis-ready basis for lightweight slum monitoring, but its applicability to slum detection - an indirectly coupled task shaped by both built form and socio-economic processes - remains untested. We evaluate AEF on slum classification and sub-pixel density estimation across 12 cities and 69 city-year pairs (2017-2024), using GRAM pseudo-masks as supervisory labels. The evaluation spans four training strategies, two protocols (random split and 3x3 spatial block cross-validation), six auxiliary feature configurations, and five baseline models, complemented by representation-level analyses (PCA, SHAP) and full-AOI mapping. Five findings emerge. (1) Same-city cross-year training is optimal under both protocols (median spatial F1 = 0.616, R^2 = 0.466); temporal expansion outperforms cross-city transfer, indicating city-scale representational drift. (2) Regression R^2 is driven primarily by zero/non-zero boundary discrimination: positive-pixel R^2 is consistently negative across all cities, revealing limited capacity to model intra-pixel density gradients at 10 m. (3) PC36 is consistently top-ranked across tasks; classification saturates at k = 32 while regression remains unsaturated at k = 64. (4) POI features yield the largest density gain (Delta R^2 = +0.064). (5) For six cities meeting dual-task usability thresholds, full-AOI inference across 2017-2024 preserves slum cluster structure (mean SSIM = 0.926). The study delineates the capabilities and complementarity needs of foundation-model embeddings for slum monitoring.

Problem

• slum detection
• density mapping
• cross-city generalisation
• global comparability
• representation learning

Innovation

• foundation model embeddings
• slum density mapping
• cross-city generalization
• sub-pixel regression
• representation learning