Mind the Gap: Bridging Prior Shift in Realistic Few-Shot Crop-Type Classification

TL;DR

This paper introduces Dirichlet Prior Augmentation (DirPA), a method to improve few-shot crop-type classification by simulating real-world label distribution skew during training, thereby enhancing model robustness.

Contribution

It proposes DirPA, a novel prior augmentation technique that models real-world label distribution skew as Dirichlet variables to improve few-shot learning in agriculture.

Findings

DirPA effectively shifts decision boundaries in classification.

It stabilizes training by acting as a dynamic feature regularizer.

DirPA improves generalization under class imbalance.

Abstract

Real-world agricultural distributions often suffer from severe class imbalance, typically following a long-tailed distribution. Labeled datasets for crop-type classification are inherently scarce and remain costly to obtain. When working with such limited data, training sets are frequently constructed to be artificially balanced -- in particular in the case of few-shot learning -- failing to reflect real-world conditions. This mismatch induces a shift between training and test label distributions, degrading real-world generalization. To address this, we propose Dirichlet Prior Augmentation (DirPA), a novel method that simulates an unknown label distribution skew of the target domain proactively during model training. Specifically, we model the real-world distribution as Dirichlet-distributed random variables, effectively performing a prior augmentation during few-shot learning. Our experiments show that DirPA successfully shifts the decision boundary and stabilizes the training process by acting as a dynamic feature regularizer.