Skip to main content
. 2022 Nov 5;35(7):5479–5499. doi: 10.1007/s00521-022-07895-x

Table 3.

Performance of the attention-aided different deep feature extraction models on the DDSM database

Attention-aided TL model’s accuracy (%)
Simulation VGG16 VGG19 ResNet50 EfficientNet
1 92.42 91.89 89.96 89.89
2 91.86 91.94 91.14 91.17
3 92.12 91.77 91.32 90.78
4 91.41 90.34 91.23 89.89
5 91.51 90.78 91.86 90.34
Mean ± SD 91.86 ± 0.42 91.34 ± 0.73 91.10 ± 0.69 90.41 ± 0.56