An Encoder-Decoder Network Based FCN Architecture for Semantic Segmentation

Table 2

Performance on PASCAL VOC2012 test set.

Method	MIoU
FCN-8s	62.2
ResSegNet	80.4.7%
RefineNet	84.2%
PSPNet	85.4%
DeepLabv3+	87.8%
Ours	85.6%