Research Article
[Retracted] Coupled Attention Framework of Convolutional Neural Network Based on Computer Intelligence
Table 3
Accuracy of ResNet methods on ImageNet.
| Method | Acc (%) | Params (M) | FLOPs (G) |
| ResNet18 | 70.34 | 11.69 | 1.82 | SE | 71.14 | 11.78 | 1.83 | CAF-SE | 71.52 | 11.78 | 1.83 | CBAM | 71.34 | 11.78 | 1.83 | CAF-CBAM | 71.56 | 11.78 | 1.83 | SGE | 69.29 | 11.69 | 1.83 | CAF-SGE | 70.30 | 11.69 | 1.83 |
| ResNet50 | 76.68 | 25.56 | 4.14 | SE | 77.18 | 28.09 | 4.15 | CAF-SE | 77.56 | 25.89 | 4.15 | CBAM | 77.63 | 28.09 | 4.15 | CAF-CBAM | 77.63 | 25.89 | 4.15 | SGE | 77.58 | 25.56 | 4.15 | CAF-SGE | 77.58 | 25.56 | 4.15 |
|
|