Fusion of Pervasive RF Data with Spatial Images via Vision Transformers for Enhanced Mapping in Smart Cities
arXiv:2508.03736v2 Announce Type: replace-cross Abstract: In this paper, we present a deep learning-based approach that integrates the DINOv2 architecture to improve building mapping by combining (possibly erroneous) maps from open-source platforms with pervasive radio frequency (RF) data collected from multiple wireless user equipments and base stations. Unlike prior methods, our approach leverages a vision transformer-based architecture to jointly process both RF and map modalities within a unified framework, effectively capturing spatial dependencies and structural priors for enhanced mappi
The increasing availability of pervasive RF data and the maturity of vision transformer architectures are enabling novel approaches to spatial mapping and urban intelligence.
This development significantly enhances the accuracy and efficiency of mapping in complex urban environments, critical for smart city infrastructure, autonomous systems, and defense applications.
Traditional mapping methods are augmented by a unified AI framework that integrates disparate data sources (RF and imagery), leading to more robust and dynamic spatial understanding.
- · Smart city developers
- · Urban planning agencies
- · Autonomous vehicle companies
- · Defense contractors
- · Traditional surveying companies
- · Legacy GIS providers
- · Manual urban data collection processes
Improved urban infrastructure management and navigation capabilities become widespread.
The integration of real-time RF data into digital twins of cities becomes a standard practice, enhancing simulation and predictive analytics.
Enhanced mapping precision contributes to more efficient resource allocation, smarter public services, and could potentially influence geopolitical strategy through superior situational awareness.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.AI