AI RESEARCH
Visual Spatial Learning: Single-Field Spatial Interpolation Using Convolutional Neural Networks
arXiv CS.LG
•
ArXi:2605.30167v1 Announce Type: cross Predicting a complete spatially correlated field from sparse observations is a fundamental challenge in spatial statistics and environmental modelling. Classical interpolation methods such as Kriging rely on Gaussian process assumptions and variography, which can limit their effectiveness in non-stationary settings and require substantial domain expertise.