Skip to main content
. 2020 Dec 2;20(23):6886. doi: 10.3390/s20236886

Table 4.

Results of training proposed model with various image sizes.

Input Image Size (pixel) MACs Parameters Average Accuracy No. of Epochs
96 × 96 6,005,248 96,408 100.00% 50
64 × 64 2,674,688 96,408 99.79% 70
32 × 32 676,352 96,408 99.58% 100