A Hierarchical Ensemble Inference Pipeline for Robust White Blood Cell Classification Under Domain Shifts

TL;DR

This paper presents a hierarchical ensemble inference pipeline that improves white blood cell classification robustness under domain shifts, leveraging a feature bank and a multi-stage kNN-based decision process.

Contribution

It introduces a memory-augmented, hierarchical ensemble approach with a DinoBloom backbone fine-tuned with LoRA for better domain shift resilience.

Findings

Ranks within top ten in WBCBench challenge by macro F1-score.

Utilizes a three-stage kNN hierarchy to enhance decision robustness.

Employs a feature bank and DinoBloom backbone for effective feature representation.

Abstract

Automated white blood cell (WBC) classification is essential for scalable leukaemia screening. However, real-world deployment is challenged by domain shifts caused by staining protocols, scanner characteristics, and inter-laboratory variability, which often degrade model performance. The White Blood Cell Classification Challenge (WBCBench) at ISBI 2026 aims to advance robust WBC recognition, with a focus on accurately identifying blast cells and other clinically critical rare subtypes. We propose a memory-augmented, hierarchical ensemble pipeline for WBC classification under domain shifts, leveraging a feature bank and a DinoBloom backbone fine-tuned with LoRA. Our three-stage inference hierarchy combines k-nearest neighbors (kNN) retrieval at each level, reducing over-reliance on any single decision. Evaluated on the WBCBench dataset, our method ranks within the top ten by macro F1-score in the final testing phase.