Table 3 Quantitative comparison (mAP) on M3FD dataset.

From: A dual-stream feature decomposition network with weight transformation for multi-modality image fusion

 

Peo

Car

Bus

Mot

Lam

Tru

mAP@[.5:.95]

IFCNN

0.546

0.698

0.82

0.51

0.648

0.667

0.648

U2Fusion

0.545

0.702

0.826

0.521

0.672

0.676

0.657

SwinFusion

0.531

0.701

0.835

0.511

0.697

0.648

0.654

SeAFusion

0.551

0.707

0.83

0.504

0.64

0.669

0.65

SuperFusion

0.545

0.696

0.815

0.509

0.648

0.683

0.649

TarDAL

0.536

0.683

0.778

0.485

0.626

0.657

0.628

SDDGAN

0.549

0.699

0.83

0.5

0.67

0.675

0.654

CoCoNet

0.491

0.662

0.8

0.462

0.613

0.599

0.604

Diff-IF

0.547

0.696

0.802

0.496

0.651

0.671

0.644

Ours

0.534

0.705

0.822

0.53

0.699

0.675

0.661

  1. The best three values in each metric are denoted in bolditalic, italic and bold, respectively.