Table 10 Computational complexity of the suggested framework and state-of-the-art models.

From: Enhancing artistic style classification through a novel ArtFusionNet framework

Model

FLOPs (G)

Params (M)

Throughput (img/s)

Inference time (ms)

ResNet-50 (CNN baseline)

4.22

25.7

221

12.4

Swin Transformer1

15.5

87.8

128

25.7

CLIP-based3

21.9

150.3

97

32.2

DeiT2

17.7

86.7

115

27.9

Proposed Model

9.8

52.4

174

18.5

Back to article page