解决血液细胞分类中的染色差异和类别不平衡问题。
Robust White Blood Cell Classification with Stain-Normalized Decoupled Learning and Ensembling
- 分步训练：先均衡采样学通用特征，再按类别采样优化分类器。
- 在挑战赛中取得第一名，显著提升罕见类型识别准确率。
- 适合医学图像分析、病理诊断等需要高鲁棒性的场景。
白细胞（WBC）分类是感染评估、白血病筛查和治疗监测等血液学应用的基础。然而，真实世界中的WBC数据集存在显著的外观差异，由染色和扫描条件变化引起，且类别严重不平衡——常见细胞类型占主导，而临床重要的罕见类型样本不足。为此，我们提出一种染色归一化、解耦训练框架：首先通过实例均衡采样学习可迁移表征，随后利用类别感知采样和混合损失（结合有效数量加权与焦点调制）重平衡分类器。推理阶段进一步通过集成多个训练好的主干网络并配合测试时增强来提升鲁棒性。该方法在ISBI 2026举办的WBCBench 2026：鲁棒白细胞分类挑战赛中取得榜首成绩。
原文摘要 · Abstract (English)
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.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。