Comparing the results of our method with ( AUTOstyle [postComp is the AUTOstyle for DSB stage 2 test] and GTstyle ) and without style transfer augmentation ( NOstyle ), we see a clear trend towards increased performance with style transfer augmentation.
← all excerpts
nucleAIzer: A Parameter-free Deep Learning Framework for Nucleus Segmentation Using Image Style Transfer.
1
—
—