Table 4 Performance evaluation for each object class on RTTS.

From: DSNet enables feature fusion and detail restoration for accurate object detection in foggy conditions

Method

Bus

Car

Bicycle

Mcycle

Person

mAP(%)

YOLOv8

0.604

0.837

0.616

0.676

0.780

70.3

AODNet

0.616

0.828

0.579

0.683

0.775

69.6

FFANet

0.621

0.840

0.595

0.679

0.783

70.3

CPAEnhancer

0.602

0.843

0.652

0.687

0.786

71.4

CDNet

0.614

0.833

0.626

0.678

0.780

70.6

CF-YOLO

0.619

0.849

0.631

0.697

0.787

71.6

DR-YOLO

0.614

0.856

0.622

0.721

0.808

72.4

RDMNet

0.46

0.75

0.28

0.47

0.69

53

TogetherNet

0.47

0.75

0.26

0.48

0.69

52.9

Ours

0.643

0.855

0.676

0.702

0.792

73.4

  1. Bold indicates the best performance.