Effective pseudo-labeling based on heatmap for unsupervised domain adaptation in cell detection

•Propose unsupervised domain adaptation for cell detection methods using a pseudo-cellposition heatmap as a pseudo label.•Introduce a Bayesian discriminator using uncertainty and the curriculum learning based on the number of cells to select confident pseudo labels.•Confirm the effective of the curriculum learning in many combinations of different domains and the Bayesian discriminator improved by using the curriculum learning.