Skip to main content
. Author manuscript; available in PMC: 2023 Jan 1.
Published in final edited form as: Pattern Recognit Lett. 2021 Dec 11;153:176–182. doi: 10.1016/j.patrec.2021.12.004

Table 1.

Comparison of Top-1 accuracy on CIFAR-10 and CIFAR-100 datasets with various baseline models and attention based classification models.

CIFAR-100 CIFAR-10
Model Top- 1 Top-1
ResNet-110 73.12 93.57
ResNet-110-SE 76.15 94.79
ResNet-110-SAOL 77.15 95.18
ResNet-110-ABN 77.15 95.09
ResNet-110-Ours 77.50 95.60
WRN-16-8 79.57 95.73
WRN-16-8-SE 80.86 96.12
WRN-16-8-ours 80.91 96.20
WRN-28-10 80.13 95.83
WRN-28-10-SAOL 80.89 96.44
WRN-28-10-ABN 81.88 96.22
WRN-28-10-ours 81.86 96.46
ResNext 81.68 96.16
ResNext-ABN 82.30 96.20
ResNext-Ours 83.02 96.43
DenseNet 77.73 95.41
DenseNet-ABN 78.37 95.83
DenseNet-SAOL 76.84 95.31
DenseNet-Ours 78.41 95.51
VGG-16 72.18 92.64
VGG-16-ours 74.67 94.29
VGG-11 68.64 92.00
VGG-11-ours 72.18 92.94