Skip to main content
. 2023 Jan 13;13:749. doi: 10.1038/s41598-023-27808-9

Table 1.

Baseline results of six different models using fully supervised learning strategy.

Model Backbone Avg. accuracy mIoU Dice
Deeplabv3 ResNet-50 96.37 48.99 0.48
ResNet-101 98.38 50.18 0.50
Mobilenet-v3 96.10 48.05 0.48
FCN ResNet-50 96.33 48.19 0.48
ResNet-101 96.38 47.83 0.47
LRASPP Mobilenet-v3 93.54 45.65 0.45

Significant values are in bold.