Semantic Segmentation under a Complex Background for Machine Vision Detection Based on Modified UPerNet with Component Analysis Modules

Table 4

Mean average precision AP on class-level of the UPerNet with/without CAM in CITYSCAPES instance-level semantic labeling task.

Method	Person (%)	Rider (%)	Car (%)	Truck (%)	Bus (%)	Train (%)	Motorcycle (%)	Bicycle (%)
UperNet	36.0	28.8	51.6	30.0	38.7	27.3	23.9	19.4
UperNet + CAM	36.0	28.8	53.0	34.3	57.0	37.5	22.3	23.8