An Uncertainty-Guided Out-of-Fold Pseudo-Label Replacement Resource for Aerial Land-Cover Data in Three Regions of South Korea
收藏官方服务：
资源简介：
"This dataset resource provides pixel-level Monte Carlo dropout uncertainty maps, map-sheet-based two-fold out-of-fold (OOF) predictions, replacement masks, and selectively refined land-cover labels derived from 12,613 high-resolution aerial imagery tiles across Gyeonggi, Gyeongsang, and Jeolla regions in South Korea."
提供机构：
Zenodo创建时间：
2026-08-01
