[2602.24251] Histopathology Image Normalization via Latent Manifold Compaction

Computer Science > Machine Learning arXiv:2602.24251 (cs) [Submitted on 27 Feb 2026] Title:Histopathology Image Normalization via Latent Manifold Compaction Authors:Xiaolong Zhang, Jianwei Zhang, Selim Sevim, Emek Demir, Ece Eksi, Xubo Song View a PDF of the paper titled Histopathology Image Normalization via Latent Manifold Compaction, by Xiaolong Zhang and 5 other authors View PDF HTML (experimental) Abstract:Batch effects arising from technical variations in histopathology staining protocols, scanners, and acquisition pipelines pose a persistent challenge for computational pathology, hindering cross-batch generalization and limiting reliable deployment of models across clinical sites. In this work, we introduce Latent Manifold Compaction (LMC), an unsupervised representation learning framework that performs image harmonization by learning batch-invariant embeddings from a single source dataset through explicit compaction of stain-induced latent manifolds. This allows LMC to generalize to target domain data unseen during training. Evaluated on three challenging public and in-house benchmarks, LMC substantially reduces batch-induced separations across multiple datasets and consistently outperforms state-of-the-art normalization methods in downstream cross-batch classification and detection tasks, enabling superior generalization. Comments: Subjects: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2602.24251 [cs.LG]   (or arXiv:2602...