AI RESEARCH
A Proxy Consistency Loss for Grounded Fusion of Earth Observation and Location Encoders
arXiv CS.AI
•
ArXi:2604.18881v1 Announce Type: cross Supervised learning with Earth observation inputs is often limited by the sparsity of high-quality labeled or in-situ measured data to use as