EURASIP Journal on Image and Video Processing

Table 2 Comparison of parameters (M), training time (m) and FLOPs (G) between our method and other segmentation networks, in which highlighted values represent the best result of each criteria

From: HR-MPF: high-resolution representation network with multi-scale progressive fusion for pulmonary nodule segmentation and classification

Methods	Parameters (M)	Training time (m)	FLOPs (G)
ENet	0.36	51	0.04
SegNet	53.54	62	4.5
PSPNet	85.86	80	5.05
DeeplabV3+	59.33	71	1.38
UNet++	9.16	42	2.16
Fast SCNN	1.2	38	0.02
DFANet	2.16	69	0.03
FANet	13.65	40	0.09
SPNet	3.8	55	0.17
Ours	3.46	70	0.16

Back to article page