SelfAdapt: Unsupervised Domain Adaptation of Cell Segmentation Models

2025/08/15 by Reith, Fabian H., Franzen, Jannik, Palli, Dinesh R. +2
#Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Machine Learning (cs.LG)

paper · doi:10.48550/arxiv.2508.11411

Abstract

Deep neural networks have become the go-to method for biomedical instance segmentation. Generalist models like Cellpose demonstrate state-of-the-art performance across diverse cellular data, though their effectiveness often degrades on domains that differ from their training data. While supervised fine-tuning can address this limitation, it requires annotated data that may not be readily available. We propose SelfAdapt, a method that enables the adaptation of pre-trained cell segmentation models without the need for labels. Our approach builds upon student-teacher augmentation consistency training, introducing L2-SP regularization and label-free stopping criteria. We evaluate our method on the LiveCell and TissueNet datasets, demonstrating relative improvements in AP0.5 of up to 29.64% over baseline Cellpose. Additionally, we show that our unsupervised adaptation can further improve models that were previously fine-tuned with supervision. We release SelfAdapt as an easy-to-use extension of the Cellpose framework. The code for our method is publicly available at https: //github.com/Kainmueller-Lab/self_adapt.

Citations

• The Entropy Enigma: Success and Failure of Entropy Minimization
• Probabilistic Domain Adaptation for Biomedical Image Segmentation
• Cellpose 2.0: how to train your own model
• Cellpose: a generalist algorithm for cellular segmentation
• FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence
• Domain Adaptive Segmentation in Volume Electron Microscopy Imaging
• A Probabilistic U-Net for Segmentation of Ambiguous Images
• Explicit Inductive Bias for Transfer Learning with Convolutional Networks
• FCNs in the Wild: Pixel-level Adversarial and Constraint-based Adaptation
• Temporal Ensembling for Semi-Supervised Learning
• Deep Reconstruction-Classification Networks for Unsupervised Domain Adaptation

Related

• CellTranspose: Few-shot Domain Adaptation for Cellular Instance Segmentation
• Cellpose 2.0: how to train your own model
• Training a universal instance segmentation network for live cell images of various cell types and imaging modalities
• COIN: Confidence Score-Guided Distillation for Annotation-Free Cell Segmentation
• CellSeg1: Robust Cell Segmentation with One Training Image