11.2. 模型重要性能数据

MobileNetv1

1x224x224x3

1.14

3793.70

0.901

0.061

Top1:

0.7061(FLOAT)

0.7026(INT8)

ImageNet

MobileNetv2

1x224x224x3

0.63

4197.62

0.767

0.090

Top1:

0.7265(FLOAT)

0.7153(INT8)

ImageNet

GoogleNet

1x224x224x3

3.00

2312.97

1.210

0.062

Top1:

0.7001(FLOAT)

0.6989(INT8)

ImageNet

Resnet18

1x224x224x3

3.65

1560.14

1.650

0.061

Top1:

0.6837(FLOAT)

0.6829(INT8)

ImageNet

EfficientNet_Lite0

1x224x224x3

0.77

2792.34

1.115

0.062

Top1:

0.7490(FLOAT)

0.7469(INT8)

ImageNet

EfficientNet_Lite1

1x240x240x3

1.20

2404.13

1.764

0.062

Top1:

0.7648(FLOAT)

0.7624(INT8)

ImageNet

EfficientNet_Lite2

1x260x260x3

1.72

2135.66

1.329

0.061

Top1:

0.7738(FLOAT)

0.7715(INT8)

ImageNet

EfficientNet_Lite3

1x280x280x3

2.77

1603.91

1.630

0.061

Top1:

0.7922(FLOAT)

0.7902(INT8)

ImageNet

EfficientNet_Lite4

1x300x300x3

5.11

1067.30

2.258

0.062

Top1:

0.8069(FLOAT)

0.8058(INT8)

ImageNet

Vargconvnet

1x224x224x3

9.06

1573.19

1.627

0.062

Top1:

0.7790(FLOAT)

0.7785(INT8)

ImageNet

Efficientnasnet_m

1x300x300x3

4.53

1141.20

2.110

0.062

Top1:

0.7973(FLOAT)

0.7916(INT8)

ImageNet

Efficientnasnet_s

1x280x280x3

1.44

2706.59

1.106

0.062

Top1:

0.7578(FLOAT)

0.7518(INT8)

ImageNet

YOLOv2_Darknet19

1x608x608x3

62.94

280.60

7.264

1.680

[IoU=0.50:0.95]=

0.2760(FLOAT)

0.2730(INT8)

COCO

YOLOv3_Darknet53

1x416x416x3

65.87

213.63

9.521

9.937

[IoU=0.50:0.95]=

0.3330(FLOAT)

0.3350(INT8)

COCO

YOLOv5x_v2.0

1x672x672x3

243.86

78.78

24.940

30.784

[IoU=0.50:0.95]=

0.4800(FLOAT)

0.4660(INT8)

COCO

Ssd_mobilenetv1

1x300x300x3

2.30

2588.11

1.102

1.101

mAP:

0.7342(FLOAT)

0.7275(INT8)

VOC

Centernet_resnet101

1x512x512x3

90.54

250.20

8.292

4.667

[IoU=0.50:0.95]=

0.3420(FLOAT)

0.3350(INT8)

COCO

YOLOv3_VargDarknet

1x416x416x3

42.82

302.11

6.866

9.921

[IoU=0.50:0.95]=

0.3350(FLOAT)

0.3270(INT8)

COCO

Deeplabv3plus_efficientnetb0

1x1024x2048x3

30.78

203.40

10.023

0.790

mIoU:

0.7630(FLOAT)

0.7568(INT8)

Cityscapes

Fastscnn_efficientnetb0

1x1024x2048x3

12.49

294.50

7.148

0.785

mIoU:

0.6997(FLOAT)

0.6928(INT8)

Cityscapes

Deeplabv3plus_efficientnetm1

1x1024x2048x3

77.05

117.35

17.009

0.771

mIoU:

0.7794(FLOAT)

0.7740(INT8)

Cityscapes

Deeplabv3plus_efficientnetm2

1x1024x2048x3

124.16

89.91

22.420

0.768

mIoU:

0.7882(FLOAT)

0.7856(INT8)

Cityscapes

Resnet50

1x224x224x3

7.72

683.52

3.144

0.090

Top1:

0.7737(FLOAT)

0.7674(INT8)

ImageNet

VargNetV2

1x224x224x3

0.72

3530.60

0.863

0.090

Top1:

0.7394(FLOAT)

0.7321(INT8)

ImageNet

Swint

1x224x224x3

8.98

138.90

14.718

0.088

Top1:

0.8024(FLOAT)

0.7947(INT8)

ImageNet

MixVarGENet

1x224x224x3

2.07

5809.93

0.646

0.089

Top1:

0.7133(FLOAT)

0.7066(INT8)

ImageNet

Fcos_efficientnetb0

1x512x512x3

5.02

1757.02

1.466

0.247

[IoU=0.50:0.95]=

0.3626(FLOAT)

0.3562(INT8)

COCO

Fcos_efficientnetb2

1x768x768x3

22.08

450.32

4.837

6.736

[IoU=0.50:0.95]=

0.4470(FLOAT)

0.4470(INT8)

COCO

Fcos_efficientnetb3

1x896x896x3

41.45

269.01

7.766

9.117

[IoU=0.50:0.95]=

0.4720(FLOAT)

0.4740(INT8)

COCO

Pointpillars_kitti_car

1x1x150000x4

66.82

117.31

32.183

2.566

APDet=

0.7731(FLOAT)

0.7676(INT8)

Kitti3d

RetinaNet_vargnetv2_fpn

1x1024x1024x3

301.27

80.92

24.689

6.540

[IoU=0.50:0.95]=

0.3151(FLOAT)

0.3129(INT8)

COCO

Yolov3_mobilenetv1

1x416x416x3

20.58

491.66

4.334

1.662

mAP:

0.7657(FLOAT)

0.7581(INT8)

VOC

Ganet_mixvargenet

1x320x800x3

10.74

2424.77

1.155

0.953

F1Score:

0.7949(FLOAT)

0.7872(INT8)

CuLane

DETR_resnet50

1x800x1333x3

202.99

47.40

41.375

1.666

[IoU=0.50:0.95]=

0.3570(FLOAT)

0.3134(INT8)

MS COCO

DETR_efficientnetb3

1x800x1333x3

67.31

62.28

32.334

1.669

[IoU=0.50:0.95]=

0.3721(FLOAT)

0.3597(INT8)

MS COCO

FCOS3D_efficientnetb0

1x512x896x3

19.94

604.57

3.994

8.708

NDS:

0.3062(FLOAT)

0.3019(INT8)

nuscenes

Centerpoint_pointpillar

300000x5

127.73

100.64

24.618

52.566

NDS:

0.5832(FLOAT)

0.5814(INT8)

nuscenes

Keypoint_efficientnetb0

1x128x128x3

0.45

3221.86

0.914

0.361

PCK(alpha=0.1):

0.9433(FLOAT)

0.9431(INT8)

carfusion

Unet_mobilenetv1

1x1024x2048x3

7.36

1050.12

2.129

0.589

mIoU:

0.6802(FLOAT)

0.6753(INT8)

Cityscapes

Pwcnet_pwcnetneck

1x384x512x6

81.71

161.49

12.671

0.304

EndPointError:

1.4117(FLOAT)

1.4075(INT8)

flyingchairs

Motr_efficientnetb3

image:

1x800x1422x3

track_query:

1x2x128x156

ref_points:

1x2x128x4

mask_query:

1x1x256x1

64.43

72.72

26.535

22.895

MOTA:

0.5802(FLOAT)

0.5776(INT8)

Mot17

Bev_lss_efficientnetb0_multitask

image:

6x256x704x3

points(0&1):

10x128x128x2

2.41

278.20

7.925

17.903

NDS:

0.3006(FLOAT)

0.3000(INT8)

MeanIOU:

0.5180(FLOAT)

0.5148(INT8)

nuscenes

Bev_gkt_mixvargenet_multitask

image:

6x512x960x3

points(0-8):

6x64x64x2

34.49

85.83

23.624

18.009

NDS:

0.2809(FLOAT)

0.2791(INT8)

MeanIOU:

0.4851(FLOAT)

0.4836(INT8)

nuscenes

Bev_ipm_efficientnetb0_multitask

image:

6x512x960x3

points:

6x128x128x2

8.83

209.61

9.739

17.980

NDS:

0.3053(FLOAT)

0.3041(INT8)

MeanIOU:

0.5146(FLOAT)

0.5099(INT8)

nuscenes

Bev_ipm_4d_efficientnetb0_multitask

image:

6x512x960x3

points:

6x128x128x2

prev_feat:

1x128x128x64

prev_point:

1x128x128x2

8.93

188.10

10.565

18.169

NDS:

0.3724(FLOAT)

0.3725(INT8)

MeanIOU:

0.5290(FLOAT)

0.5388(INT8)

nuscenes

Detr3d_efficientnetb3_nuscenes

coords(0-3):

6x4x256x2

image:

6x512x1408x3

masks:

1x4x256x24

37.55

27.04

69.306

2.410

NDS:

0.3304(FLOAT)

0.3283(INT8)

nuscenes

Petr_efficientnetb3_nuscenes

image:

6x512x1408x3

pos_embed:

1x96x44x256

36.24

8.41

226.049

2.418

NDS:

0.3760(FLOAT)

0.3733(INT8)

nuscenes

Centerpoint_mixvargnet_multitask

300000x5

51.45

106.09

23.532

50.427

NDS:

0.5809(FLOAT)

0.5762(INT8)

MeanIOU:

0.9129(FLOAT)

0.9122(INT8)

nuscenes

Stereonetplus_mixvargenet

2x544x960x3

24.29

244.52

6.407

15.433

EPE:

1.1270(FLOAT)

1.1352(INT8)

SceneFlow

Densetnt_vectornet

goals_2d:

30x1x2048x2

goals_2d_mask:

30x1x2048x1

instance_mask:

30x1x96x1

lane_feat:

30x9x64x11

traj_feat:

30x19x32x9

0.42

86.58

26.587

10.518

minFDA:

1.2974(FLOAT)

1.3038(INT8)

Argoverse 1