Segmentation and Classification of White Blood Cells Using the UNet

Table 2

Comparison of results sets of existing supervised methods.

Methods of architecture	Mean accuracy	IoU	B.F score	Precision	Recall	Specificity	F1 score
UNet	93.4	90.2	0.65	92.55	97.12	92.74	94.50
SegNet	92.14	85.6	0.52	98.77	97.66	99.89	99.10
FCN	91.34	92.6	0.72	95.65	96.77	97.45	98.67
Proposed method	94.14	95.6	0.92	98.45	97.56	93.23	98.67