Table 3 Comparison results of different models in the experiment.

From: Small fishing boat classification and recognition method based on GASF temporal graph coding and EMPViT model

Model route

Model type

Param.

GFLOPs

ACC

CNN

VGG-16

6.3

1.3

84.37%

ResNet50

5.8

1.6

86.34%

GoogleNet

6.5

2.1

86.89%

DenseNet

7.1

1.8

87.62%

MoblileNet

8.2

2.3

88.72%

Transformer

ViT

6.0

3.3

92.26%

Swin

7.3

2.7

94.75%

LocalViT

6.2

2.5

93.58%

DeepViT

6.8

2.9

94.66%

CaiT

5.1

2.1

95.28%

CrossViT

5.7

2.4

95.81%

MPViT

5.4

1.6

96.79%

EMPViT-Base

4.3

2.9

98.84%

Back to article page