Robust White Blood Cell Classification with Stain-Normalized Decoupled Learning and Ensembling

TL;DR

This paper introduces a robust white blood cell classification method that uses stain normalization, decoupled learning, and ensembling to handle appearance variations and class imbalance, achieving top results in a challenge.

Contribution

The proposed framework combines stain normalization, decoupled training, class-aware rebalancing, and ensembling to improve WBC classification robustness under real-world conditions.

Findings

Achieved top rank in WBC classification challenge

Effectively handled appearance variations and class imbalance

Enhanced robustness through test-time ensembling

Abstract

White blood cell (WBC) classification is fundamental for hematology applications such as infection assessment, leukemia screening, and treatment monitoring. However, real-world WBC datasets present substantial appearance variations caused by staining and scanning conditions, as well as severe class imbalance in which common cell types dominate while rare but clinically important categories are underrepresented. To address these challenges, we propose a stain-normalized, decoupled training framework that first learns transferable representations using instance-balanced sampling, and then rebalances the classifier with class-aware sampling and a hybrid loss combining effective-number weighting and focal modulation. In inference stage, we further enhance robustness by ensembling various trained backbones with test-time augmentation. Our approach achieved the top rank on the leaderboard of the WBCBench 2026: Robust White Blood Cell Classification Challenge at ISBI 2026.