Learning Multidimensional Urban Poverty Representation with Satellite Imagery

2025-09-05

AI Summary

Existing urban poverty inference models rely solely on urbanization features, failing to capture spatial inequality and exhibiting weak correlation with ground-truth poverty indicators. To address this, we propose a multidimensional representation learning framework that jointly encodes accessibility, morphological, and economic features extracted from high-resolution satellite imagery. Crucially, we introduce a counterfactual adjustment mechanism grounded in building morphology to mitigate spurious correlations inherent in nighttime light (NTL) data. Methodologically, our approach integrates contrastive learning, building footprint extraction, NTL intensity modeling, and causal backdoor adjustment for robust multimodal fusion. Experiments across Cape Town, Dhaka, and Phnom Penh demonstrate substantial improvements in informal settlement poverty identification accuracy—particularly in data-scarce regions. Our framework delivers interpretable, transferable, and policy-actionable insights for precision poverty alleviation.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Recent advances in deep learning have enabled the inference of urban socioeconomic characteristics from satellite imagery. However, models relying solely on urbanization traits often show weak correlations with poverty indicators, as unplanned urban growth can obscure economic disparities and spatial inequalities. To address this limitation, we introduce a novel representation learning framework that captures multidimensional deprivation-related traits from very high-resolution satellite imagery for precise urban poverty mapping. Our approach integrates three complementary traits: (1) accessibility traits, learned via contrastive learning to encode proximity to essential infrastructure; (2) morphological traits, derived from building footprints to reflect housing conditions in informal settlements; and (3) economic traits, inferred from nightlight intensity as a proxy for economic activity. To mitigate spurious correlations - such as those from non-residential nightlight sources that misrepresent poverty conditions - we incorporate a backdoor adjustment mechanism that leverages morphological traits during training of the economic module. By fusing these complementary features into a unified representation, our framework captures the complex nature of poverty, which often diverges from economic development trends. Evaluations across three capital cities - Cape Town, Dhaka, and Phnom Penh - show that our model significantly outperforms existing baselines, offering a robust tool for poverty mapping and policy support in data-scarce regions.

Problem

• Inferring urban poverty from satellite imagery using multidimensional traits
• Addressing weak correlations between urbanization and poverty indicators
• Mitigating spurious correlations in economic trait inference

Innovation

• Integrates accessibility, morphological, and economic traits
• Uses contrastive learning for infrastructure proximity encoding
• Incorporates backdoor adjustment to mitigate spurious correlations