BMC Bioinformatics

Table 2 Summary of optimal training parameters determined using each network for each training set

From: Application of convolutional neural networks towards nuclei segmentation in localization-based super-resolution fluorescence microscopy images

DataSet	Optimal parameters	Network
DataSet	Optimal parameters	Mask R-CNN	Stardist	ANCIS
Tissue	Epochs	400	400	200 (500 RPN)
	Steps	500	300	NA
	Learning Rate	1E−4	1E−4	1E−4
	Training Set	77*	77*	77*
Cell	Epochs	600	400	400 (500 RPN)
	Steps	500	300	NA
	Learning Rate	1E−4	1E−4	1E−4
	Training Set	60	65*	65*

^*Indicates the number is the maximum number of images available in the dataset

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com