图片

方法

OA

IoU

P

R

F1

图片1

TransUNet

0.9171

0.9033

0.6792

0.9024

0.7751

SwinUNet

0.9306

0.9217

0.8159

0.7256

0.7681

U-Net-vgg16

0.9508

0.8159

0.8867

0.7905

0.8358

U-Net-Res50

0.9221

0.9117

0.8532

0.6711

0.7513

DeeplabV3+

0.9504

0.9400

0.7993

0.9172

0.8542

CUT_UNet

0.9387

0.9297

0.8026

0.8129

0.8077

MSANet

0.9641

0.9580

0.9210

0.8617

0.8904

图片2

TransUNet

0.9246

0.8968

0.8637

0.8914

0.8773

SwinUNet

0.9319

0.9088

0.9321

0.8357

0.8812

U-Net-vgg16

0.9401

0.9197

0.9575

0.8392

0.8944

U-Net-Res50

0.9254

0.9013

0.9383

0.8063

0.8673

DeeplabV3+

0.9505

0.9322

0.9392

0.8944

0.9162

CUT_UNet

0.9487

0.9297

0.9328

0.8949

0.9135

MSANet

0.9594

0.9441

0.9575

0.9061

0.9311

图片3

TransUNet

0.8675

0.7517

0.9068

0.8147

0.8582

SwinUNet

0.8880

0.7763

0.9223

0.8307

0.8741

U-Net-vgg16

0.8764

0.7574

0.9267

0.8057

0.8619

U-Net-Res50

0.8884

0.7752

0.8842

0.9170

0.9003

DeeplabV3+

0.8995

0.7925

0.8851

0.9390

0.9113

CUT_UNet

0.7070

0.7851

0.9128

0.8488

0.8796

MSANet

0.9238

0.8446

0.9334

0.9274

0.9304