Table 1 Performance results on ImageNet for EfficientNet variants and different variants.

From: A fine tuned EfficientNet-B0 convolutional neural network for accurate and efficient classification of apple leaf diseases

Model

Top-1 Acc.

#Params

#FLOPs (Approx)

EfficientNet-B0

77.1%

5.3 M

0.39 B

EfficientNet-B1

79.1%

7.8 M

0.69 B

EfficientNet-B2

80.1%

9.2 M

1.00 B

EfficientNet-B3

81.6%

12 M

1.80 B

EfficientNet-B4

82.9%

19 M

4.20 B

EfficientNet-B5

83.6%

30 M

10.3 B

EfficientNet-B6

84.0%

43 M

19.0 B

EfficientNet-B7

84.3%

66 M

37.0 B

ResNet-50

76.0%

26 M

4.1 B

DenseNet-169

76.2%

14 M

3.5 B

ResNet-152

77.8%

60 M

11 B

DenseNet-264

77.9%

34 M

6.0 B

Inception-v3

78.8%

24 M

5.7 B

Xception

79.0%

23 M

8.4 B

Inception-v4

80.0%

48 M

13 B

Inception-resnet-v2

80.1%

56 M

13 B

ResNeXt-101

80.9%

84 M

32 B

PolyNet

81.3%

92 M

35 B

SENet

82.7%

146 M

42 B

NASNet-A

82.7%

89 M

24 B

AmoebaNet-A

82.8%

87 M

23 B

PNASNet

82.9%

86 M

23 B

AmoebaNet-C

83.5%

155 M

41 B

GPipe

84.3%

557 M

-