Mohamed-Slim Alouini, Fares Fourati, Zakarya Elmimouni
We have not lifted any functions out of this paper's repositories yet, so there is nothing we have run. If it links a repository, it is listed below.
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
Accurate school detection is essential for supporting education initiatives, including infrastructure planning and expanding internet connectivity to underserved areas. However, many regions around the world face challenges due to outdated, incomplete, or unavailable official records. Manual mapping efforts, while valuable, are labor-intensive and lack scalability across large geographic areas. To address this, we propose a weakly supervised framework for school detection from aerial imagery that minimizes the need for human annotations while supporting global mapping efforts. Our method is specifically designed for low-data regimes, where manual annotations are extremely scarce. We introduce an automatic labeling pipeline that leverages sparse location points and semantic segmentation to generate infrastructure masks from which we generate bounding boxes. Using these automatically labeled images, we train our detectors on a first training stage to learn a representation of what schools look like, then using a small set of manually labeled images, we fine-tune the previously trained models on this clean dataset. This two stage training pipeline enables large-scale and strong detection in low-data setting of school infrastructure with minimal supervision. Our results demonstrate strong object detection performance, particularly in the lowdata regime, where the models achieve promising results using only 50 manually labeled images, significantly reducing the need for costly annotations. This framework supports education and connectivity initiatives worldwide by providing an efficient, extensible approach to mapping schools from space. All models, training code, and auto-labeled data will be publicly released to foster future research and real-world impact. Impact Statement-School infrastructure mapping is a key component of global education and connectivity initiatives. However, many regions still rely on incomplete or outdated records, while manual annotation of aerial imagery remains costly and difficult to scale. This work introduces a weakly supervised detection framework that overcomes these limitations by combining automatic labeling from sparse geolocation data with a two-stage training strategy. The proposed approach significantly reduces the need for human annotations while maintaining strong object detection performance, enabling scalable deployment across large geographic areas. By improving the efficiency of representation learning under limited supervision, this framework can support large-scale mapping efforts such as the Giga Initiative. It is applicable to infrastructure monitoring, urban planning, and public policy, and can facilitate better resource allocation in underserved regions, contributing to more inclusive access to education and digital services worldwide.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2605.03968")
get_code_for_paper("2605.03968")
have("2605.03968")
Connect an agent — have() is free.