Table 4.
Evaluation comparison results for blood cell identification via the proposed BCNet against the latest deep learning works in the literature.
| Reference | Data | Methods | Az. (%) |
|---|---|---|---|
| Zhao et al., 2017 [44] | Cell vision, ALL-IDB, Jiashan | CNN, SVM, and random forest | 92.80 |
| Journal et al., 2021 [66] | Collected, BCCD data set | Two DCNN | 95.17 (Precession) |
| Acevedo et al., 2019 [32] | Private | CNN + Transfer learning | 96 |
| Qin et al., 2018 [29] | Private | CNN | 76.84 |
| Ma et al., 2020 [40] | BCCD | DCGAN + Transfer learning | 91.7 |
| Baydilli and Atila 2020, [67] | LISC | Capsule network | 96.86 |
| Rui Liu et al., 2022 [37] | HPBC | Transfer Learning | 96.83 |
| The proposed BCNet | HPBC images | BCNet | 98.51 |