Skip to main content
. 2023 Oct 30;9:e1665. doi: 10.7717/peerj-cs.1665

Table 3. Comparison of network performance on three datasets.

Model Flops (G) Imagewoof The Oxford-IIIT pet 102 Category flower
Top-1 Acc Top-5 Acc Top-1 Acc Top-5 Acc Top-1 Acc Top-5 Acc
PVTv2_B1 2.1 85.4% 97.5% 79.7% 93.5% 95.6% 98.7%
Swinv1_Tiny 4.5 78.8% 97.7% 70.7% 93.7% 94.1% 98.4%
CMT_S 4.0 79.7% 96.7% 59.0% 82.3% 74.9% 87.5%
MViTv2_Tiny 4.7 85.6% 97.2% 83.2% 96.0% 95.3% 98.9%
Swinv2_Tiny 6.6 75.8% 97.6% 66.4% 91.2% 91.1% 97.6%
Ours 5.7 87.6% 98.7% 82.6% 96.8% 95.4% 98.8%

Note:

The results of the model presented in this article are shown in bold.