Table 8. Rank-1 Accuracy comparison of proposed FV-EffResNet with other methods using SDUMLA dataset.
| Train/Test split | Method | Accuracy (%) |
|---|---|---|
| Training 3 Testing 3 | Gumusbas et al. (2019) | 87.00 |
| Chai et al. (2022) | 96.61 | |
| FV-EffResNet | 95.75 | |
| Training 4 Testing 2 | Das et al. (2018) | 97.48 |
| Gumusbas et al. (2019) | 88.00 | |
| Boucherit et al. (2022) | 89.88 | |
| FV-EffResNet | 98.40 | |
| Training 5 Testing 1 | Gumusbas et al. (2019) | 100.0 |
| Boucherit et al. (2022) | 99.48 | |
| Huang & Guo (2020) | 99.53 | |
| Liu et al. (2022) | 98.11 | |
| Ma, Wang & Hu (2023) | 99.82 | |
| FV-EffResNet | 98.45 |