GeoCFM: Positive-Only Conditional Flow Matching for Mineral Occurrence Sampling

2026-09-28

AI Summary

This study addresses the localization challenges in critical mineral discovery arising from positive-only samples, the absence of reliable negative examples, and ambiguous geological features by formulating target area prediction as conditional spatial distribution learning. Methodologically, it discards pseudo-negative strategies and instead leverages a UNet to extract multi-channel geophysical image features. Combined with a point-conditioned velocity prediction mechanism, conditional flow matching is employed to directly learn a two-dimensional pointwise transport field, effectively bridging dense grids with sparse supervision for mineralized point set generation. Experiments on both synthetic and real-world USGS datasets demonstrate that the proposed approach significantly improves the geometric consistency between predictions and observed mineral occurrences compared to baseline models, while effectively characterizing epistemic uncertainty.

Abstract

Critical mineral discovery is a positive-only problem: deposits are observed as sparse locations, while unlabeled regions are not reliable negatives, and similar geophysical signatures can arise from different subsurface states. We therefore model mineral targeting as learning a conditional spatial distribution over occurrence locations, $π(p\mid d)$, given geo-images $d$, rather than predicting a deterministic per-pixel score map. We introduce GeoCFM, a conditional flow-matching model that generates mineral occurrence point sets conditioned on multi-channel geo-images; GeoCFM learns a point-wise transport field in $\mathbb{R}^2$, using UNet features with point-conditioned velocity prediction to bridge dense rasters and sparse supervision without pseudo-negatives. On a synthetic magnetics--geochemistry benchmark with latent activation and on USGS Earth MRI data with a spatially disjoint tile split, GeoCFM improves geometric agreement with observed occurrences over score-map and non-conditional baselines, while representing epistemic uncertainty through conditional sampling.

Problem

• Critical mineral discovery
• Positive-only learning
• Conditional flow matching
• Mineral occurrence sampling
• Epistemic uncertainty

Innovation

• Conditional Flow Matching
• Positive-Only Learning
• Mineral Occurrence Sampling
• Point-wise Transport Field
• Epistemic Uncertainty